 Can you please take a seat? So good morning and welcome to this last panel discussion because before we move to the really important people after lunch and We're going to come back to the issues which which we've discussed yesterday and this morning and very much hope that We can also connect What we've heard this morning on the on the micro aspects of price formation and wage formation and what we've learned yesterday on the on the on the macro big picture and so you're we're perfectly allowed to come back to Questions on the on the Phillips curve. It's okay can come back in the discussion and It's pretty much going to be about the same issues that we've already discussed but I would very much like to Encourage the panelists and urge urge the panelists to Try as much as they can to distill the policy consequences Because that's that's what what we're here to discuss both at the micro level and at a macro level And also I would like to encourage the panelists as much as they can to focus on the European dimension of that discussion including possibly on the eurozone dimension of that discussion given the sequence we're in in terms of Decisions on the in the future of the eurozone and there is a lot to say about the importance of labor markets and wages also in in Eurozone adjustment So if that dimension can come up in the discussion that can be very useful for us as as policy makers I'm going to be tough on time, but we have a good incentive structure. I guess because Next item is a lunch And so all the stakeholders here Have an interest in disciplining the panel So that we don't steal too much time out of the lunch so Let's let's start the initial presentation Erica Welcome the floor is yours Great. Well, thank you to the organizers certainly for including me here today And it's a pleasure to be back in the central banking community where I spent most of my professional life But as many of you know that for the last four years Until 2017 I was head of the Bureau of Labor So when I want to talk about today is central bankers and national statisticians and I want to talk About some of the things that I learned as Being head of the BLS that I think are very relevant to central bankers and You see this picture here of a Cristiano Ronaldo confronting problems with imperfect statistics Butting up his head there. I think that this He may not know that that's what was frustrating him, but we do So I'm going to go From there and talk about some price index measurement challenges and then talk about improvement opportunity Opportunities and challenges and I think of them very much as part of central bank policy Okay So let me start off by saying well, what is the the natural the appropriate relationship between Central bankers and national statisticians well There's certainly the client and provider relationship that we think of immediately that central banks are very important providers rather clients of The national statistical system, but of course they're also very much professional peers that The expertise the Professionalism of both central banks and national statistical offices makes them a natural Colleagues in dealing with issues They have very compatible Missions first of all their overall mission to broadly support national welfare is very consistent But their particular missions are quite distinct They really should be as As president Draghi reminded us yesterday They really need to be independent of each other in order to preserve their credibility That's very important But this independence the independence is important both in reality and because of optics But this independence shouldn't mean abdicating support for each other and That I think is an important issue so let me I'll return to that in a minute Let me talk a little bit about the goals of a national of a statistical system agency And I remember I put here in the trenches because the important thing to realize is just as Central banking has a nitty-gritty aspect to it and imperfections involved in what it does That's very much the case for statistical offices So the goals are obviously the best possible official statistics to guide decisions. So why do we have statistical? Agencies sometimes those of us who are data nerds forget why it is we love statistics But it is because it empowers us to make good decisions and it powers everybody in the economy to make good decisions Not just central bankers and what are what's good data five characteristics accurate objective relevant timely accessible Okay, I can go into those in great detail, but I think you all understand what they mean This though has to be subject to very real constraints resource constraints Putting out the data within a time frame so that it can be timely So when we're talking about the CPI from collection to production 20 days, right? And you have to do that every month You have to preserve respondent confidentiality you have to avoid undue burden so that you keep response rates up and Because of that you make changes only when you're certain that they will improve accurately Significantly So that's the world in which the statistical agencies operate so when a Economic phenomena is noticed comes up a very important question to ask is is this real or is it a data artifact and Certainly no national statistician would argue against asking that question when when the phenomena occurs But there is a slight difference in approach to this because the central bankers and the researchers will think of this often as a new question something they haven't run into before and They'll say oh this really demands attention and resources and let's throw them at the question Where's the statistical agencies by and large have been dealing with these questions for Since they created the statistics in issue at issue, right? So very often these issues that are brought up are an acknowledged limitation of statistics because it's all statistics have Limitations and have been the subject of ongoing improvements of research and institutional memory So there's a little bit of difference one group saying wow and the other group saying again That's the world in which they deal with Each other but joining then joining forces can be amazingly productive because the national statisticians bring in this institutional memory this experience is understanding and The the researchers the central bankers have this energy and this focus on it. So With that in mind, let's talk about price index measurement challenges I'm going to focus to on two issues. It just came up recently That have been raised recently one is new goods and quality adjustment bias and the other substitution So we're talking about innovation in price indexes. I'm really going to run through this, you know the basic the basic Approach is to try and match products over time and to that So the the more closely you can match products over time The more you can attribute any price change to inflation. That's the basic idea And this gives you an idea from one year of the the CPI that In that year only two percent that yellow piece there of the prices in the CPI were Where cases where quality adjustments were needed so by and large price matching really works But we do need to do quality adjustments some of the time There's here three ways that quality adjustments are done producer provided quality adjustment That's more on the PPI and import export prices input from other surveys is another source and then there are hedonic adjustments CPI uses a lot of that and housing is a very big part of that so But some quality adjustment and new goods bias remains BLS is always kind of playing catch up on all of this and a recent paper that Four co-authors and I put out Yeah, in 2017 April 2017 in the JEP tries to estimate how much that is we come up with an estimate of a bias on real GDP of 4 tenths of a percentage point from this bias We find that There's been a little change in this over time. It's real. It's there a little change over time It's mostly not about the digital economy. It's mostly about health care Okay, and it looms larger when obviously when growth is slow then then this is a larger part of what you're seeing From the BLS and BEA perspective Nobody there was actually really alarmed at this, but neither were they satisfied Okay, it has helped to focus Efforts, so if you're interested in this I would Point you to that paper All right substitution bias. This has also been a longtime concern on average It causes overestimation of cost of living increases so to the extent that you want to Measure cost of living changes you get that There have been ongoing improvements in the CPI to deal with this and then there's been the introduction of the change CPI to really try and address the Substitution bias and that has a monthly changing Market basket and it also uses the torn fist Formula to allow for more subs to control for substitution bias So just to give you an idea of whether that is getting worse or better You can compare the CPI to the chained CPI and this blue line is the is Difference in the 12 month. It's a rolling Average difference in the 12 month CPI and chain CPI from the inception of the chain CPI 2000 up to 2017 and then I've drawn in the red line just You know divided the sample in half and said what's what's the difference and if anything the substitution bias seems to be decreasing So We push right ahead because I am running out of time oops. I've already run out of time Okay, so improvement opportunities and challenges many opportunities There's progress being made This gives you an example of places where progress is being made within the BLS and be a Opportunities The main opportunities that I want to call your attention to our uses of alternative data And they're really quite a few Alternative data sources that are being tapped within for the statistical agencies I list here some of them, but I think it's very important to realize that none of the sources is free Riskless or clean So it actually takes resources to begin to tap into these things Research is required comparisons because these data are so important to so many decisions. You can't go about Carelessly change Okay, there are also some very important interesting issues that can be looked at The role of margins in inflation I think is very interesting the new producer price index system calls out and and provides estimates on margins particularly the the PPI trade services margins I Often got complaints about what was this thing doing? It's large. It's volatile and I think the answer is a group of Bankers and or researchers can probably now that there's a series long enough can probably now begin to understand what these margins are doing And so I encourage somebody to to take that up Another question is work hours. We don't measure and understand work hours very well I think that there's a lot of room for Attacking and understanding work hours and then a big very a big topic is That I think has been under under researched and That the statistical agencies should Be funded to delve into more carefully is are our wage indexes in the US. We have the the average hourly earnings which has a For which there's a lot of item non-response so it's smoothed a lot It's not as good as you might want it to be and there's the employment cost index, which is Composition adjusted. It's a good measure, but it's only quarterly There are a lot of opportunities. I think for improving our compensation not just wages, but compensation measures Now I'm really over. Okay, so conclusion. Let's see. Okay All right, so we have many improvement challenges The main thing I want to say here is that there is a role for the central banks to support their statistical agencies That's very important What kind of support do I mean I mean support we now have We have statistical agencies under attack for their integrity Central banks need to speak up for the integrity of the central of the statistical agencies This is happening certainly in Puerto Rico and in the US, but also in Greece where the former central banker is has been convicted of the crime of publishing accurate GDP statistics And many other countries where the statistical agencies are underfunded And then of course they're standing up for the integrity of the data. That's hugely important. So let me stop now because I'm over Thank you. Thank you Thank you very much Erica We can come back to any issue in the in the discussion will have plenty of time to To discuss and on the in the case of Greece I would like to by the way to confirm that the that upholding the indebtedness and integrity of indebted agencies and in particular of the statistical Production process in Greece has been a priority of the European institutions and remains a priority and That will also be on the table In the overall discussion tomorrow Philip the floor is yours Hello everybody, I'm pleased to be with you. So I am from the ILO international labour organization We don't used to be invited in there A central bank meeting. I'm not an academic. Well, I asked twice. Are you sure? Okay, so I will come back on some of the issue we we listen this morning in particular maybe About more general trends and I will have only four slides because I will try to stick to my 10 minutes So the first is figures that we already know The first figure that has something to do with what we we see this morning compares the evolution of labour productivity and wages in advance economy, huh? So the data are From 1999 until 2017 for the group of the 36 developed economies Here instead of hourly wages figure The paper presented in this morning. We use the more widely available mostly wages and consistently or measure of productivity in GDP per worker We also deflate as it has been done this morning by the CPI and GDP deflator and we observe the same tendencies that with the GDP deflator The gap with labour productivity is smaller, but in any case The conclusion is the same we we see a decoupling between wage growth and productivity growth Among the country we also see that this is not a recent phenomenon. It is a long-term trend that has lead to the decline of labour share in a majority of Countries in the world in a majority of developed economy too In this chart index real wage ex-labor productivity are weighted average So this means that big country influence the figure more than small countries and The figure is very driven by US, Japan and Japan In reality, of course, you have a diversity of situation across countries We see we listen this morning about France. You can say the same with Italy where you have a Productivity growth and wages going together, but productivity is not increasing. This is what we observe now I move to the third to a second slide. Voila, okay We as it has been say I think it is very important not to approach which did I make only by looking at average But also the distribution the centil and the size of sir which distribution experienced different evolution a long time and Or calculation and OECD calculation Right to the same conclusions that wage inequality has increased in Develop economy obviously this has important implication because as we know I Wage earner of the higher propensity to save the unmiddle class and low pay worker So if average wage cross only from the an increase in the top distribution The effect on aggregate demand and eventually on price might be Less than if it is a both broad based wage increase that benefit middle-class and low paid workers To illustrate the extent of wage inequality in Europe, so I tried to stick to Europe We produce these 3d figures when we rank enterprise in the oriental axis According to the average wage We rank the workers according to the way annually early wage in the in-depth axis and on the vertical Axis we rank we Illustrate the level of the wage we are speaking about and what we see in this in these figures that you have in Europe a lot of workers that are paid less than 20 Euro per hour and These are mostly not in every countries, but mostly the workers for which which has stagnant stagnated the part on the Upper right side corner of the figure I tended to increase more Some recent study have emphasized the need to reduce inequality between enterprise but our figure shows that inequality within enterprise is also very important and According to our calculation close to 80% of workers earn less than the average wage in the enterprise in which they work and Within wage inequality count for 42% of total wage inequality in Europe So it's not ridiculous. This means that reducing wage inequality within enterprise Require is part of the policy response to wage inequality and the possible impact now. I will move toward to labor market institution. I Will look now First the contractual arrangement so over over the past decade Part-time work and fixed-term contract have been increasing in Europe part of these increases conjonctural Will use the phase figure of Spain for example, but we see that there is a structural tendency to have an increase of This form of employment and we expect that technology technology can change Will further this if I diversify employment pattern in the future. There is a range of legal instrument Set equal treatment and equal pay for workers in all form of contractual arrangement including international labor standard and EU directive but in practice as we see in the graph temporary employment controlling by the characteristic of the worker and characteristic of the job are suffering wage penalty sometime part-time employment Feature premium, but that is rarely the case in Europe in Europe also part-time employment is associated With wage penalty and I think we'll listen something about that this morning working worker in temporary employment multi-party employment relationship and out-call work have Little participation in trade unions are less member and more broadly they're bargaining power tend to be Lower than standard workers for multiple reason including fear of retaliation Sometime diverging interest with standard workers lack of our best of right so and This is very controversial There is also from part of the trade union movement the feelings that non-standard worker argues Not only to lower their bargaining power as them, but to lower the bargaining power of trade unions But I will not enter in that you know So non-standard form of employment can help enterprise to adapt to the demand in the labor market contributing to sustainability of enterprise and gross But what so we are not for to get rid of this form of employment But we want in the ILO to ensure that this job are good job Where workers have earnings that are predictable and in line with equal pay for work of equal value Now I will look at another labor market institution trade union and collective bargaining that have also a strong influence on wages as we listen this morning, but as we know This shows the evolution of trade union density and collective bargaining coverage in Europe from 2016 So it's quite amazing because we see that only two country experience an increase in collective bargaining coverage During this period for EU 28 the decrease is of more than 20 percent So significant and all country with very few exception experienced the decline in trade union density from more than 10 percent with extreme cases such as Lotia, Estonia or Slovakia where the decline is of more than 50 percent so it is Despite the general discourse that they used to listen in the European Commission for example about how important is social dialogue and collective bargaining we observe an erosion of trade union density and collective bargaining bargaining coverage in Europe that is relatively strong and I would say that Often what we observe it was not the case in in Germany not too much a case in Germany But often public policies are at the origin of the declining coverage of collective bargaining So to conclude So there are different reasons and I will listen some about real wage stagnation Sometimes it's only a problem of low productivity growth or sometimes is a problem of decoupling and We I love we focus on the evolution of labor market institution in this area that we can the bargaining power of workers because We suspect we didn't quantify yet that this is application in term of wage level I just want to conclude by saying that I am convinced that the decline in the collective bargaining coverage that we observe is not an inevitable consequence of the changing world of work I think history has demonstrated that collective bargaining has a flexibility and adaptability required to face change for me, but it's up to the social partner to say It's a big challenge to have modality of employer and worker organization and of collective bargaining that are more inclusive among a diversity of working Workers in different form of employment Including the new forms that emerge in the digital economy I will also say that collective bargaining to be strong need good public policy or favorable public policy and I think there is a lack steel of Research to understand better our different collective bargaining arrangement Can contribute to both equity and performance in the labor market? Thank you very much Philip. I mean your slide on the decline of Union coverage and collective bargaining is absolutely striking I guess one obvious question you can store it and answer later But one obvious question is how the change relates to the level Focusing on France for instance the change was limited, but it starts from from a very low level So when when other way to ask to ask the same question is do we see convergence across Europe? In terms of for industrial relations or or divergence which matters also for the functioning of the of the region But that's for later. Tomas or the floor is yours Thank you very much and thank you for having me on this Panel it's a pleasure to be here I'm I'm an IO economist. I'm currently working for the European a commission competition watchdog So I want to give you a sort of antitrust a perspective on markets Competition authorities what they do they try to prevent prices from going up So I understand in this conference a puzzle is what is why we don't see Inflation going up. So I think I saw the puzzle we can go for lunch. It's because of the job We do I wish it was that that it was that simple so Being more serious now, I will give you some ideas about concentration matters The debate started a couple of years ago in the US. So this is but I will present European data So in the US there is some evidence that the market concentration is going up What this figure is doing is looking at the change in a concentration index HHI and in some industries This change is very substantial. So in 80% of US industries concentration has been going up between 1997 and 2014 So that's a that's a large change and have a picture that You've seen is on mark cups and margins This is from the paper of yanna echoed and yanna a deloc the lurker you see that from the 80s Something changed and margin. So the ability of firms to price above cost There is a big debate on what is the notion of cost, but the ability of firms to price Above cost has increased margins are on the rise. So what's happening instead in Europe first? Thing to say is that some companies are actually global So whatever is valid for the US would be valid for Europe to because we belong to the same global market But instead we started some so a problem of Europe is that there is less Systematic data collection. We are trying to improve. So at the Commission. We have an ongoing project by the end of a summer We will have a time series of 20 25 years There's just to give you a flavor of it is just over six years Concentration in Europe has been pretty Stable so not lots of dynamics here. You see the five largest economies in the EU plus the aggregate you the the the The picture is rather static at the industry level some regularities start So there is huge differences across industries a regularity here Is that you will see that they really concentrated industries in the according to all different measures are telecoms transport manufacturing and finance Within each industry there will be at the native because when I say ITC there will be fixed telecom mobile telecom Semiconductors computer so it will be different things. We have to be a bit more precise Another thing as you start digging deeper into industries You will see that in Europe many of these markets are still largely national. So so we don't really have a single market We don't have a lot of integration So it's really telecoms in Italy are telecoms in Italy telecoms in Spain are telecoms in Spain and so forth So we don't have a single market Margins that's really where something valid maybe across the board starts emerging So this is our own data on European margins And they are compared with the US and you see apart from a striking differences after the recession and Europe It took much longer, but but but if you look at the trends, they are actually quite similar So margins are on the rise. So in a sense a takeaway of some of this Literature is that no matter what a method you employ So yeah, and for instance is is doing data at the firm level and estimate production function This is instead from national accounts. These people using accounting data in different ways. This seems to be a regularity So this seems to be a regularity having said that Don't be too excited This is the European picture of the five largest economists and the picture is driven by Germany France and the UK to some extent Italy and Spain now they are recovering margin But they really suffered after the a financial crisis and margins dropped to very low levels But they are picking picking up again So what are the implications for us for you because we have slightly different perspectives on this The first implication as an economist I would say if concentration goes up you would expect prices to go up and quantities to go down Chetris parables all as equal then we can discuss whether everything else is indeed equal So these are two things prices going up, but also quantities So when you measure your inflation, etc having fixed baskets may may may be capturing only one aspect and not the old picture So quantities do matter on every implication for And a crowd like this one is that to the extent of market power is a Substantial feature of the aggregate economy it needs to be factored in in our models And I'm sure you're already doing this So I'm not familiar with the way you do macro models But just think of my thoughts as something that maybe you're already doing that's fine But if not try to see how you may have to change a few things so account for markups But these markups are not constant over time They're not constant across industries to have seem kind of simple monopolis in competition models may not be exactly the way To go you may have to adjust it. I said that this is the the the prediction the simplest microeconomics Perspective prices on the increase in quantities on the decrease Chetris parables So what may have changed and this is interesting about margins going up everywhere is that there seems to be a common Substantial technological change going on so the cost function of firms seems to have changed or more economies of scale Say to simplify higher fixed cost and lower marginal cost These implies that it's easy to show that you will have less entry the few firms in the markets We'll have higher markups and prices may or may not go down according to how the The market power is compensated by cheaper cost cheaper marginal cost The the the consequence of this is also that will be less entry So one of the big war is from a policy a perspective we have evidence from the US We start having evidence in Europe every year firms, you know a born and firms die So these entry and exit dynamics are super important on the exit that dynamics We are more or less on the same historical trends of firms die now as they used to die in the past Instead firms don't seem to be to born as much as in the past. So we don't have this Dynamics of this healthy dynamics that we we had in the in the past So some implication also for you and this I'm repeating something that already I've said And the the pass-through is related to market power So think about the effectiveness of some of your monetary aid transmission mechanisms Under some conditions going from competition high pass through to to less competition should reduce the past through So this is a general theme that the the responsiveness of firms changes This means that money type Transmission will be affected but in general responses to shocks will be will be affected can be Responsed to shocks in in the labor markets could be Responsed to engine exits. So we might have less dynamism question mark. I don't know. This is really a conjecture You should also think sometimes maybe you're already doing that in your models of the vertical structure There's upstream and downstream in the kind of concentrations. I'm observing at the commission There's lots of boring stuff but essential stuff So market power being acquired over cement over titanium dioxide or raw materials and you name it But these will be used by a lot of industries and if these industry themselves have market power What you will expect on the final prices would be would be would be different. This is not a general result though Aviva mentioned I think tobacco and tobacco said this is an industry where the Past through is higher than a hundred percent I would say it's a more general feature of some industries where say a mast have goods Industries where demand is very convex. We don't have to be technical But the price goes up people still want to buy pretty much as much as what they were buying before so when demand is very convex actually The more a market power you have the higher the past through will be so when you're doing your studies Think about the composition of industries. There's lots of heterogeneity there in most sectors consumer goods. I would Conjecture Monetary policy to be less effective because what I said the past through bill will be will be lower But you know these essential goods not only tobacco, but can be transport energy They actually actually the other way around could also be a possibility And finally I have my last minute when I want to talk a little bit about digitization So gaffer this Google Amazon Facebook Apple. So These are three of these companies will be Reaching one trillion dollars of market capitalization soon. So do they have any an impact in the economy? Yes, of course But not as large as you might expect to in aggregate We heard this morning that all the digital stuff Erika said it's being measured which is great. There is Just to give you a Perspective is it market power going to increase the prices or is the super efficiency of this? Superstar firms are going to compensate for it So using a typical merger simulation model we use at the commission to simplify if you have quite a big Concentration from three to two so three major player to two under some assumption Typically at the the prices would increase substantially by 10% if costs were equal and because they acquire market power They decrease in marginal costs need to be more than one for one The decrease in marginal cost would need to be of the range of 16% So just as an order that just serves a favor a flavor of what's going on Are these kind of efficiencies what we're observing in the data that could justify the facts that prices don't go up despite the market power? Finally a debate which is more on the Google side. We are paying with our data. So we because Search is free or Facebook. We do we don't need to pay a Subscription a subscription fee because then these guys get to know a lot about us and then they do targeted advertising So are we not measuring prices, right? Yeah, I would say prima facie. We are already measuring prices in the sense that all this Information is it then used to do targeted advertising these feeds into the production function in the cost function of those firms Which are interested to advertise online. So to the extent that these costs are already Imputed in our measures then we should not do double counting because we are already to the extent that indeed the advertising costs are Something that you monitor. I'm gonna finish here. Thank you very much The last speaker is Klaus Zimmerman and we are hoping back from product markets to labor markets. Yes. Thank you. Thank you very much. I Have no slide so you can see Also, I think that it's much easier to interact with the policy issues when you have no structure Opposed on yourself. So you have more flexible now But we learned what we have to do in a panel is is to find out what what is the policy Implication of what we have heard so far on the day So we have somehow to review what what goes on and I will certainly focus more on Vages and on prices now we found out as one Finding is that Shopping is exciting even without inflation. We also learned that nice unions Do it all as least at least as long as they exist and These are some provocative Positions already now when I wrote 40 years ago my diplomat thesis on the rational Expectations and the phillips curves well inflation was money yes and Well, the phillips curve wasn't really existing anymore. Yes because of rational expectations Nowadays, it's completely back. We are back. So it speaks the phillips curve exists and well money is not really relevant Yes, at least I haven't heard much about money the two days, but still I mean a Simple formula of that time is is helpful also explained by labor economists are here. We are here We have a reason to be here Well, if the labor share is constant what many of you are doubting Yes, then inflation is nothing more than the difference between the wage growth and labor productivity growth. Yes, that's a simple tautological reformulation so well wage growth and labor productivity and these are labor issues Yes, it's only this issue of that using as a labor share, which is maybe not fully labor but okay, so we in anyway have a reason to be here and Central banks can learn I guess something from the work we are doing also on the micro level now The first issue now is why are we concerned because when I was invited was are we are concerned about wage Vages wages are not rising enough. Yes, it needs Yens The white man in Germany to push for higher wages not the unions Yes, Yens white man for years pushes for higher wages in Germany. So really this is a concern Why is this really a concern now? Well wages are prices and prices have signals and Signals are important even if wages would not have to rise. Yes Some of the wages in general but some of the wages have to go up others have to go down to signal changes in structural developments in the economy to relocate and well, but you also know from research of Economists and one of them one of those is here even well one of those is this problem, Julie Who had written years ago a book and later all the articles on it? Why? Why is there are no cut wage cuts? Yes wages are never go down wages are all this go up If what stays the same and this has a lot of psychological reasons So if you want to see signals wages have to grow because some Vages can grow a little bit more other can grow a little bit less Yes, so to to provide the information since already one arguments are many other arguments Why wages should go up, but that's at least from the start something what is very important very central to the market Economy now the question is of course, what what are wages? Yes, well This is the same debate on what are prices? Yes Somehow labor costs or labor labor income is divided to some kind of human index Yes, some some some some take heads others take hours But well if you think about labor hoarding Robots and and all kinds of changes working over time unpaid lay unpaid work and all these kind of issues It's not not an easy concept So looking at wages and to see why some countries behave different or I'm regions behave different some sector sectors They behave different. It's very important to see the institutional backgrounds to understand what's what's behind and now Yeah, the question is of course what determines wage growth and we are a little bit puzzled and we learned not only here but also here that see See Phillips curve whether with inflation on the left-hand side or with this wage growth at the left-hand side is Flattening at the moment and substantially flattening and says puzzling about it. And the question is this falling unemployment rates The issue is the challenges issue build unemployment can unemployment falls so low Let's say in Germany or in the UK or in the United States that if prices would have to to explode and well, we had first we have to realize and that's the work we have seen by By by stock and by Schoenberg yesterday and today is that there's this evidence for flattening of the Of the Phillips curve there's also work by David Bell and David plan show recently very fresh new papers They are arguing that the issue with the unemployment rate has very much to do until after the recession with the way How unemployment measures the underutilization of labor? so what we don't measure is that people would like to work more or Previously would have liked to work less but no longer do so. So they show that people working Are not working enough in their own opinion and others who in the past have thought they work tomorrow Too much I work wanted to work No longer want to work less. So there is an underutilization of work And if you if you bring this into a theory in this regression, you'll find much more stable results also the general finding They show is that we are still flattening however, they also point at an interesting debate between Cains and and beverage Who well beverage already in 1944 was predicting a low unemployment rate and Cains was very Untraceable that this could be could happening that 3% is very very low now in 1960 purely reflected backwards arguing well I was not so wrong and by the way between 1948 and 1959 in the United Kingdom the average Unemployment rate was 1.5% Without in prices exploding. So the idea is maybe we are should not be so much concerned about That issue that challenge In in general now, of course many factors determining wage growth and explain differences across countries There are measures Scarcity is of course an important issue productivity changes Of course price expectations, but no longer at the moment We are now back to past inflation as a predictor of unemployment Expected price expectations. This was was once defeated 40 years ago as wrong And we have of course gross expectations driving the issues future labor markets Well, if Rafa robots will really do what we expect or some people expect from them Then we will have large stocks of labor in the future under employed at work if they cannot be fired under employed at work Which is a prediction then that wages out in the future will be low However, but I also haven't heard much in this conference so far is migration is helping also to influence not only wage changes, but also to Make wages adjust smoother across areas There is research done by saying that in particular in the country of the European of the euro in the euro zone the adjustments across countries to Asymmetric jobs have have have increased Which is what we want and this implicitly of course says also that wages would would would smooth Down, okay now the last point I would like to comment on is on Germany rightly we have not heard a nice paper and nice discussion of that paper Which also concentrates on what my Germany is different and I agree with many of the particular of course the empirical facts are very well Elaborated the question is interpretation the question is interpretation I'm a little bit more doubtful that this was all done by reunions who like To save the German economy if one has maybe part of the of the heavy debate Ongoing about what is right or wrong one can have different opinions on this One then one has to smell that must be more on it So so for instance is the is the internal is true. That's a huge. It's a huge internal flexibility Which have been achieved and others including myself and co-authors have also worked on that issue, but without a long-term pressure on reforms and changes These kind of changes would not have been taken Place also several lots of government intervention. So the labor hoarding In the great recession by the organized by the German government was also a measure of internal flexibility so Also, it's true that more and more companies no longer follow the union wages, however Theoretically it's possible every ministry is a federal ministry, but also the state ministry for labor have a commission where one of the social partners can apply for a Generalization of the wage contract. It's true that it is no longer done so much and the question is why? Yes, I mean says there is a debate. So so also unions behave in a way Which are also the German unions. Yes, I have to follow interest. It's in the long term They have to also think about wages. They decided to to to bargain for for jobs, but that is I think a second issue so so the issue of zero forms Could be debated more and I stopped here, of course because my time is over but I think the reform policies in Germany directly and indirectly by affecting expectations and put pressure on on on all partners in society have been important so I would argue just recommending the German model to France or Spain or Italy by saying the unions have to be more constructive It's I think not Export good. We which will be successful. That would be my predictions. Thank you very much Thank you very much Klaus. Let me give a chance to the panelists to to react to what we've heard before I turn to the floor if you want to is there anything you want to say? No No questions. Okay. Good. So let's go to the questions. Let's let's open the discussion I will take the questions three by three And Starting with Richard's protest and please introduce yourself nevertheless a question for Erica you Referred to you scraped over very quickly the question of scraped data. Yes, and I'm a little bit familiar with the billion prices project which seems to be a rather impressive. Yes effort and I'd be interested in your comments. I mean you said not free riskless or clean I think Well, yes, but okay. I mean maybe you can Give us a little more detail on on the comparison between that and US CP CPI data and because there are periods if you look at the at the indices There are periods in which there is significant divergence But over time they catch one one catches up with the other And and do those different periods of difference tell us anything Thank You Richard Maybe one additional question to Erica if I may when it comes to web scraping and and big data generally is about Representativeness of the other samples because there might be an illusion that when you have very big data It has to be representative, you know, it's like in this Jorge Luis Borges story where the ultimate map is the country itself, right? So if you get all data is representative But big data is also subject to sample bias adverse selection and the like so that would like to know more about it Luigi Zingales Yeah, Louis Zingales from Chicago. This is a question mostly for Tomaso. In fact is a point because I'm sure you know that But you dismiss Like a joke the tension between macro economists and I or macro economists say no problem because prices are Too low that don't rise and then the problem by your economists have been solved now as you know, and then you point it out in your data Are your economy is concerned about margins not about price levels? Macro economists concerned about price levels Why do I say that because I've seen people dismissing for example concern about the telecommunication industry in the United States because if you graph The prices the price has been going down and concentration has been going up. So say there is no problem Reality if you look at how much prices drop they drop much most slowly than in other countries And I think that's the reason the concentration generated that but then there is an other interesting question It is more for the macro economists and actually for the ECB the central bankers is we know that The reaction to price increase and decrease is very different So if you look at gasoline whenever there's a price drop of gasoline The gasoline pumps are less prompt in rebating it price increase is more So to what extent the inflation targets Impacts actually the IO of pricing of companies to the point that favours tacit collusion in a moment in which you have Reduction in cost due to innovation and then the fact you have a low inflation targets allow Firms to keep prices constant in face of dropping costs and so favours collusion Thank You Luigi. Let me take a third question Here please and sorry if I can't identify you, but just too far away central bankers only have a medium-term horizon We're not that far-sighted. Thank you central bunker is here. I must get out there from from Bank of Spain My question is to to Tomasso as well And it's in connection with the the relation that you established between markups and the past through off of monetary policy you essentially said that at some point that the fact that Markups are now higher may imply that the past rule of of monetary policy may be maybe lower I must say that they found the argument a little bit Unclear for in principle you may have a high level of markups But these markups are still being very Responsive to to to some shocks including possibly to to monetary policy shock So my my first question is a clarifying one I wonder whether you can allow rate a little bit more on on this because it may be an important issue for central bankers Given the evolution that you show us about about markups now I Think that there is another channel that that may affect the past rule of monetary policy and the effectiveness of the of the measures Deployed by by central bankers these days, which is the one linking higher markups Profits retain profits and the financing of investment by by no financial Corporates now What we have seen over the crisis is that in some cases this was true in the US Markups have been strongly Contracyclical and in the euro area markups were counter cyclical precisely or more Contracyclical precisely in those countries that were more severely hit by the financial crisis actually in countries like Spain Portugal or Ireland markups went up in the worst part of the of the crisis and some of my colleagues have documented that this this Phenomenon was particularly intense in those firms that were more financially constrained So what we are seeing these days is that markups are going up that firms are able to finance an increasing part of their financing needs through retained profits my impression is that this may eventually erode the effectiveness of of monetary policy think about for instance German firms German firms have been recording a Substantially positive net financing capacity. They are net lenders to the rest of the of the system in a relatively large Amount and we have seen this before in countries like Japan and we are seeing these in countries like like Spain these days as well So essentially this is telling us that firms are not investing much given Very very loose financing conditions. They are pilling up a lot of cash. So in this context Perhaps we should expect that the small changes in the interest rates Engineer by central bankers perhaps are not going to have a big effect But I would certainly like to know your views about this about this issue. Thank you Okay, thank you. Let me turn back to to Erica and Tomasso Let me remind you that there is a trade-off between the length of the questions and the number of questions For the next one's every careful, okay So so that's a really interesting question of the comparison between the CPI and the billion prices project There are two things to general points that I can elaborate on the first of all that official statistics are Transparent in a way that will always be inconsistent with protecting profit making intellectual property. So there they're Billion prices project as As good as it is We'll always have some proprietary methods that that you will never be able to examine and know Because they have to protect their secret sauce The second thing is that there are data limitations from web scraping that That billion prices solves by relying on official statistics So they use the consumer expenditure survey for their waiting They use CPI numbers for all of the products and services that are not Posted on the web. All right And they use the CPI for validation and for benchmarking. So they are very much an extrapolation interpolation in some ways of the official CPI a Complement to it useful. They put out some tailors and products that are tailored to special needs that That people are willing to pay for but they are not a substitute for the CPI And now in terms of web scraping in general, well, it depends on what you scrape and what you scrape for This CPI program does use web scraping for actually for its For many of its hedonic models now because that's the cheapest way to download information on product characteristics these days And so that's that's very useful. It is experimenting and thinking about how to to scrape prices on an ongoing basis, but the problem with that is that twofold one is Informed consent issues right now if you participate in the survey You know you participate in the survey and you have promises from the BLS on how that information will be treated the legal issues for Consent for web scraped information haven't been worked out and I'll tell you Confidentially a census in BLS have different advice from lawyers on on this so it hasn't been worked out and along with that is that There are high fixed cost to setting up a statistical process to Produce dependable data if you build your Process around something that might disappear that you have no control over then Then all of a sudden you may not be able to produce data of reliable quality when that happens So so there's some some big barriers generally. I see these big data projects as useful for the statistical system and obviously for the people who pay for those products But they are a complement not a substitute for official statistics Thank you for the questions When you are a Unofficial of the commission you develop a skill in deflecting The question that you answer on a slightly tangential way, so I'll try this skill if I have acquired it You'll never be as good as central bunkers So Luigi yes, absolutely But but I was thinking of that trends so I wasn't comparing levels so trends are common Some trends are common across us in Europe The fact that some some some prices are higher in the US. I do agree I'm really worried when the when the head of the Antitrust division of the Department of Justice is saying that there is no evidence that a fourth Mobile firm in markets reduces prices or anticipating that the next merger way will be approved Luckily in Europe. I think we are slightly more independent Maybe by design because we oversee 28 was soon 27 different countries So the system of revolving doors is not the same as there So but I would be more concerned with for about your bill that than your your your bundle in the US Expectation about inflation so I do observe a lot especially in so-called declining industries or these basic stuff So for instance recently we had the merger among Tao producers It's a disgusting product how is the thing you made this the cigarette filter with and cigarettes are declining because Worldwide apart from a few countries going down And so they said you should allow us to merge from from three to two because we'll be able to rationalize capacity And we said this is not the right counterfactual So if sure you will have to rationalize capacity anyway But but but if you want to increase prices of cigarettes for health reasons We have different tools and they the merger was actually blocked So but we see a lot of this that that sort of implicit collusion going on especially in declining industries of homogenous goods on On the question the flexion I guess the My the purpose of what I said was to throw some ideas into the modeling of Micro-economist as I said most likely you are already are doing it if you don't perhaps think about it to to to Clarify the the link is not between margins and a pass through between between Concentration and pass-throughs if concentration increases path through are predicted to go down But if concentration increases also margins so indirectly there is also this kind of Correlation what's interesting for me since I am on the anthracite side is to see the effect to which monetary policy What you do affects what I'm doing. So since I'm observing an unprecedented at least in the last 20 years Wave of M&A's of mergers and acquisition I wonder whether the monetary policy to the extent that it keeps it gives a cheap credit to firms Especially to very large firms these firms want to acquire other firms instead of doing organic growth Which is what we would prefer from a company competition point of view question mark I would be I would like to hear your views in fact and and also another mechanism I've been interested in Again, if you provide cheap credit, you may actually Induce less of this industry shakeouts that we had in the past So in Europe did we keep too many inefficient firms alive? I don't have an answer to that, but I would like to know to what a tank monetary policy can can affect that Thank you. Let me take another round of questions. So we had John Milbauer Thanks John mobile And Tommaso Valletti made some very interesting points and they link with the discussion this morning from from Aviv Nebo on Concentration I Think you said to my sir that you assume that macroeconomists would have taken concentration into account in their inflation models Actually, I think that would be the wrong assumption and I think part of the reason why they haven't is because they've been focused on Inflation models where the price level doesn't matter because concentration and markups affect relative prices So if we think about inflation as partly a process of relative price adjustment, then Concentration and markups become an important part of the dynamics and The work that you're doing at the Commission in terms of providing data on concentration at the national level going Back 25 years. I think it's really valuable in improving inflation models and the grueland paper that you cited For the US for the first time gives us a comprehensive Huffing doll index going back to 1970 or 72 and actually Janine R and I And I've been modeling us inflation and it turns out this is actually a Useful part of the model. So it affects relative prices. So union density So worker power and firm power are part of the process of relative Price determination and that in turn affects inflation dynamics Thank you seven em cadence yours come and then we'll have Charles whiplosh and there will be a last round of questions Okay, sorry. I'm Shepnam kalem nose cam from University of Maryland. My question is for Tomasso I want to Compare I want to ask you in terms of the work you are doing for Europe How it is different than the US because I want to compare us and Europe in this whole markup business But we know by now is the the markup increase in US is driven by large firms In fact, if you look at the census firms You don't you don't find much and the new work done at the fund for European firms Pretty much coming with the same conclusion last year in Sintra We also see the differences between Europe and and us Okay, this says if it's about large firms We should be thinking about antitrust and mergers and acquisitions and all those things now coming back to Europe Part of that is also integration within Europe, right? like the foreign mergers and acquisitions across the border and that's something that is desirable and Wanted and you know supported by the policy since the introduction of euro because that's going to bring risk-sharing equity financing and all that So my question is this work you are doing, you know looking at the increase in markups large firms all firms What did you see any role for foreign firms? You know when we have more you know foreign acquisitions across the border German and friends if that like top four firms Driving the ratios are they all for and what will be the implications for that because that's something I believe we want in your opinion I just wanted to to to react to to John will power I Don't understand the reasoning that we when we think about inflation we should care about relative prices or markups or all of that If they have if these things change They change once and the change can last for two years five years ten years But you cannot imagine that markups keep going up or keep going down because they would hit the limit You cannot think that concentration will keep moving in such large amounts so at the end these are temporary effect and the subtle banks that care about the medium run are Very unlikely to see any impact of these things and I'm afraid there is a lot of confusion building up or unless maybe I am Confused but I think this is something and I think Ricardo raise raise the same issue earlier on So we spend enormous amount of time discussing these things But at the end of the day the impact of inflation may be beyond two three four five years is likely to be negligible So to put it differently that does any of what we've been discussing over these two days matter from my entire policy That's a pole back to the back to the panel Mother you want to start? Yes. So thank you John. Thank you for for this comment I think no matter what your your opinions are it's great that we're discussing about it So having so it's true also also before somebody said we're typically not invited in this kind of for I think it It's excellent that ECB in the past. It's collecting data. I know there is a An effort done by ECB to collect concentration indices and the fact that we are here together is can can just be a positive thing The what we're doing in particular the commission is is twofold. We are gathering information about firms in Plausible anti-trust markets because for me the market definition is an important exercise so where products compete So it's not so three-digit level industry would be to aggregate for me so we are doing something to do a market reconstruction exercise, which is not an obvious one and And then we are doing also a collection or a systematic collection on the on the margins of the firms that we observe Which is a very selected sample because obviously the commission is looking at super large kind of kind of murder So by a Monsanto a hundred billion dollars This is not the average firm But we do observe that margin for those large firms is is on the rise historically We know less about the the smaller firms Erica, yeah, I would say also that the theoretical foundations of many of our models assume a market structure that's that's more or less Competitive and if we are going to a world where that's not Generally the case then we should expect some changes in the behavior of Variables that we that we care about because we have actors who are no longer in a competitive market and so I think knowing what's going on in the In how markets operate it's got to be very important for conducting monetary policy because it gives you an idea of where to look and What sort of relationships might be changing? Thank you, Klaus. Yeah, that's coming back Charles to your question What can we say from the discussing so one example is minimum wages where I think The long-term effects are very important also for pricing Because that's in Germany. We had no minimum wages because the trade unions for many years didn't want minimum wages because they were controlling the wages Anyway, now since they're no longer can it can't control it. That's why they've stuck I was in meetings where the Regulants that we need the minimum wage now because you have no other way to influence wages and this is not only for the poor It's all for everybody because if you if you push wages minimum wages high Then all the upper wages also hopefully move up. So it's about a long-term strategy to move up wages You may be in favor of it against it. I'm not touching this at the moment I'm just saying so now the Commission in Germany if you have is is making proposals every so so to speak years Also to increase the minimum wage and by that lead the debate about how the wage Structure should should develop. So that's that's something here, but at least has to be observed by by central banks. I would argue Okay One of one of the interesting arguments made is that if you let the minimum wage sink low enough then it can become a tool of Of employers it becomes essentially a monopsynistic tool They can all set their implicitly or explicitly decide to set their wages to it and then Yeah, it becomes a way of collusion Providing a method for collusion When there isn't an obvious other mechanism for it and in that case then raising the minimum wage You could get actually increase in wages and increase of employment at the same time And if you didn't understand that as being a part of a labor market You could be quite confused by watching wages go up and employment go up at the same time Thank you, maybe in line to what you say for example in Europe in UK. You see that is a wage profile in the design is Far down by minimum wage because basically employers use minimum wage to pay the minimum they can So you are the cure of the of the median what is the distribution of wage distribution in the UK is like this, you know So this is a tool that can be in double effect positive or negative for workers So let me take a last round of questions. I already have three questions. So let's let's consider the list is closed So starting with Ricardo Ricardo Reis So with the start of the of the banking and monetary union in Europe What we observe was an enormously large capital flows from the court to the periphery and it was this is in between 2000-2005 and 6 and what we observed as well Which was I think quite surprising was that this coincided then with a complete stagnation of productivity in Portugal Greece Ireland many countries And that has led many macroeconomists who have used IO tools and say that there's a there was an enormous amount of Misallocation of these very large capital flows from the court to the periphery and to see in that a beginning of where the euro crisis Started and many of the problems that then are still afflicting us and let some of the discussions earlier now a difficult in social For instance has worked quite a bit of this and so have I A difficult that we've had in macro and so for the IO economists exactly where could you help us is well We observe the markets. We observe the concentration. We've gone and measured who have used the IO tools Well, we have difficulties understanding. What's the source of misallocation? From the perspective of capital not service of market power, but from capital Is it that financial systems in Portugal or Greece weren't mature enough to allocate this enormous influx of capital that came from the core? Is it that banks weren't sufficiently regulated or benefited from too low interest rates or lax monetary policy? Is it that governments intervened to protect some sectors and some companies and some incumbents were powerful or is it that antitrust? Antitrust authorities didn't do their job or not What way can you enlighten us in this debate and use maybe the IO tools a linking the macro capital flows? to the To the misallocations as opposed to the more micro concentration approach. It's a luigi or you were mentioning her Okay, one question here Yeah Yeah, Nick out to University College London I just wanted to pick up on the point that Charles made I Agree that if this is temporary the increase in the in the markets Then it shouldn't matter too much for monetary policy the issues or the facts are that you know It has been going on for 40 years and the increase in the markets are by about 40 to 50 percentage points So we're going from 1.2 mark up to 1.6 1.7 And it's not just that we see it there So that's a that's 1% a year in additional inflation if you want if that would be the economy aggregate But we see it in the decline in the labor share Which is consistent with that because it's basically the reverse of the same effect And we see it also something that hasn't been talked about much in terms of firm profits in the US firm profits At least for the publicly traded firms has gone from 1% of sales to 8% of sales over that same period now again That's consistent with this increase in Market power, but the point I'm trying to make is that this is not something that's just very short Okay This has been going on for 40 years and the pickup has been going on until now in fact for the last six seven years The increase has been the most steep and then the last point is it's the same story in Europe At least we look at the publicly traded firms and in response to the Shaznam's point the European firms have the same Story there's an increase of about 40 points in markups and also in in profits Thank you Can you pass the mic to Michael for the last question because I have a really quick question for Erica I'm fascinated by first-degree price discrimination. That's becoming sort of rampant on the internet And whenever I go on the Expedia I have my wife do at the same time so we can see we get the same price offer So how does this how does the BLS? Even dream of tracking that I mean this could be changing over time. I mean, I think These firms are getting more and more important and I think Amazon is probably also practicing some type of First-degree price discrimination, you know my credit card. They know how much money I make in print or how much I spend How do you deal with that? So actually I actually don't know that in particular. I mean, I know how they price I actually watch them price utility costs and and Telephone services using the internet and in that case they signed on as if they were a customer and put in their request and Solved what the price would be so I assume that that's what's going on as well So whatever those cookies are reading these characteristics as Is probably getting the price, but I don't know that anybody's looked at it And I would think that that would be quite an interesting a project Very quickly on actually on first-degree price discrimination to the extent that it exists I just want to make the abstract but but but interesting point that pricing would be based on willingness to pay So just on demand side consideration and cost are completely disconnected So monetary policy is lost forever in the in that kind of world because the prices don't react to cost Which is just before lunch food food for thought is always interesting And and to Ricardo again, we can continue this discussion obviously But in terms of misalocation one I've already said to the extent to that investors follow profits They want to make money Instead of organic growth where you have to compete against someone else So you're just gonna acquire companies. There is evidence of these mergers which are killing some particular companies So they all especially in the in the in the biofarm and industry they all the adagio was The small guys should be bought by the large guys because they don't have enough capital And the law is gonna be supplied by the large guys in the data We observe that's true to the extent that they are not competing against each other So the research projects are completely disconnected if instead the research projects are very close to each other as you might expect from I know your point of view the small guys are being bought and then the project is discontinued which produces a Misalocation another misalocation could be in terms of inputs I said think of the other vertical structure market power is on the input So they may be there the wrong choice or the inputs that then also Trickles down to the final prices. We have a a consumer surplus Benchmark in our module analysis, but I can see some of these mechanism But again, thanks a lot for your observations. Just one note of caution tomato monetize policies Not is not only even not even not primarily. I would say acting on cost It's about acting on aggregate demand, right? So just one note of hope for for for us. Absolutely Thank you. Close Philip one last word One word of occlusion Yeah inequality was an issue which we have also covered which is Relevant I think that what we have seen in many countries including Germany that the brighter spread In the wages It's not necessarily an inequality issue because we have to conceses in in the context of households and have Income households in total have that's one important aspect. And also what I mean Michael already mentioned that and also with a Schoenberg that that if Low skilled people get a job They have low wages and this doesn't mean that it's not worth that they can get these jobs I think that's very relevant to understand Well, I think what one of the things that we didn't touch so much that is important with inequalities inequality between groups women man my grand national and I think this is an important dimension in terms of policy response in The way our next global which we report will be exactly about the composing gender Inequality in term of wages that they and for example in the last report It's interesting to see is that the general gender gap controlling by the characteristic of woman and job is around 20% in Europe In the first top 1% it is 45 45% or 47% Thank you for this last remark. Thank you very much to the panelist. This was a fascinating discussion I get the only conclusion is that it's about the beauty of general equilibrium Know that we it takes labor economists and competition and IO economists to say something meaning meaningful on prices So that's a conclusion. Thank you very much