Symposium Day 3. Resource Misallocation and Manufacturing Productivity: The Case of Ukraine

they i would like to present a paper which is titled to resource misallocation and manufacturing productivity the case of Ukraine the structure of my presentation is introduction leader Xi methodology data has an equilibrated pyramid parameters empirical results and finally I’m going to conclusions so while looking at relation between business conditions in a country and its GDP it’s possible to notice this graph as the x-axis that’s doing business conditions so the right better to the left hours and on the y-axis that’s GDP per capita so one can notice that the better the better doing business conditions in the country the richer country however as due to the shape of the scatterplot I I concluded that minor changes are not working so only significant improvement of doing business conditions can boost growth of countries so this the these conclusions so from simple graph motivated me to understand how by improving doing business we can improve our output so and how the mechanism works to within the methodology I am applying my marking environment effects productivity in a country and productivity effects output and this idea came from different from different area so it first it’s mackerel at macro level in Simon’s seminal papers in gross here it was proved to that tfp is one of the main sources of economic growth in a country then it it was proved that actually it is the most source of growth some time later when micro foundation of macroeconomic started to emerges this flow it it was just it was proved that it wasn’t cratic distortions which affect each country they each firm the effect their productivity so is better to study overall productivity from micro level perspective and finally the methodology which I wish I implying it assumes that market conditions they they affect firm firm level productivity because because of some distortions firms are unable to employ optional optimal amount of resources so they are less productive as a result productivity in the industry is also lower and this lover in overall output and thus the country becomes were so that’s my motivation is using this methodology to to provide rationale for reforming doing business environment and research question which I important in the in this paper is what are the potential gains of optimal resources allocation in Ukrainian economy and studying learning literature on this question I decided to impose initial hypothesis that better allocation of resources can improve productivity in Ukraine Ukrainian manufacturing by from Sochi 282 a eighty percent and this literature emerged from paper by psy and and clean of which into science at night nine published their paper where they studied effects of micro level resource misallocation on manufacturing productivity in India and China and after that a number of studies was published and this paper is is a part of this flow over off literature and the contribution of this paper is to perform this exercise for Ukrainian economy so speaking about methodology general stop is that I i applied see and click and cleaner model we which is based on closed economy

variation of of malice model that’s monopolistic competition model visited hitter gymnosperms and firms differs in productivity and distortions they face and also it should be mentioned that coefficients in production functions they were I between industries but not between every specific firm so at the firm level goose are produced Swiss cobb-douglas production functions then while we aggregate them into industry level output C is egg product production functions is applied and finally single final goods which has produced by a presented firms it combines outputs of manufacturing industries using again cobb-douglas production function and as I as I already mentioned that firms here differs differ by distortions let’s face a paper by D’s markets enrichment least we’re good examples of such distortions which which are covered by this methodology so for example what what can distort films production for example non-competitive banking system provide fiber favorable loans to do some producers based on non-economic criteria on the other hand financial institutions can can be unable or unwilling to provide credit to firms that are highly productive but they have no credit history oh they can’t prove that they can pay back but they are really good on the other hand government may may offer subsidies may offer special tags deals all lucrative contracts to specific producers on the other hand various product and labor market regulations also can change cause of labor and a cost of capital enforcement of tax collection also can distort firms production and finally finally of course its corruption as corruption in a grain is very important factor factor which distort the production of Ukrainian firms so but in order to calculate these gains of optimal resource allocation I follows three main steps of it first i compute firm level tfp then i moved to industrial level gfp and finally I aggregated in aggregate manufacturing productivity so at firm level as it was mentioned at previous as a previous presentation who who heard it so we compute output wage and capital wage output wage is is defined as relationship between between share of let of labor which should be applied according to production function coefficients and we compare them to actual a share of labor and the output which is shifts both capital and labor productivity in one direction on the other hand capital wage is computed as relation between share between relation between labor and capital as it is as it is predicted by production functions and the actual the actual relation based on these distortions we compute marginal revenues as i say i was told that these distortions they affect productivity of factors of capital so for example in the marginal revenue of labor we used we use output wedge which distorts optimal wage and for marginal revenue product of capital we adjust optimal rental rate by these distortions then we aggregate these into total total physical productivity which is actual productivity of firms a firm and total factor revenue productivity which which is believed to decrease when firms are productive and and and increase when firms are less productive as more productive forms can impose lower prices so the revenue productivity will be lower and less productive firms have

enough room for maneuver to to decrease prices in order to be profitable so this total factor revenue productivity is the math is the main measure of of distortions at firm level then we move to industry level at the industry level we be aggregate productivity of every single form to to average 22 average industrial level productivity and we aggregate it as harmonic average of physical revenue which is actual physical productivity which is actual productivity of firms and we wait is by by deviation of revenue productivity of firm from the average revenue productivity in in the industry s as i was told that in ideal case they should be they should they should be equal to the mean then in a deal if effective situation a effective industry total total factor productivity will be just the mean physical productivity and at aggregate level we again just aggregate these industry level productivities to the aggregate productivity and aggregate output and gains of of liberalisation of market conditions which means that we rule out all the distortions are computed as real as as relation between actual total factor productivity overall productivity and the effective one and and then gain we just we just are calculated in in in percentage terms and here I i would like to mention that this is a formula for an ideal case and so all the distortions are ruled out however this this situation does not seems seem to be real as in any case some distortions will will affect industry in the country so we also apply a benchmark benchmark distribution of of distortions and I would like to apply this the same page benchmark which see and Cleon of applies that the US economy so I take the measure of the T of a tfp gains and adjust Ukrainian tiff begins by these gains and obtain potential chief begins if we move to us distribution of distortions speaking about data just briefly speaking I used balances and financial statements of all manufacturing companies in Ukraine in the period is two thousand from thousand to two thousand ten i use revenue wage bill employment book value fixed capital stock material cost industry code ownership exported and important indicators i can i construct a value edit fixed capital measure accidental variable and age variable I I adjust data for deflator ‘he’s for revenue fixed capital materials cost i adjust by production but by producer price index wage bill is adjusted by CPI some steps for data cleaning are made in order to satisfy that all firms have positive revenues and positive inputs i dream those cases where production function estimation was was wasn’t satisfied the properties of a cop data production function also as it is as it is done in the seminal paper which which which i replicate I tree the outliers here are tales of TF p & veggies distribution so the most productive the least productive firms are are dropped and also the those firms which which are subject to the highest distortions as a result we I came to final data set set of almost 218 observations for approximately fifty-three thousand to Phoenix firms and the price the price of data set moral more or less robust to outliers that from year to year I lost

from 20 23 to 32 percent of value added some distributed some descriptive statistics are presented here so the cips descriptive statistics for the whole sample inflation just adjusted data over over the over the last period so we can we could see that turnover well a edit away wedge bill material material costs increase to 2009 which was the year of financial crisis in Ukraine so it was a dropped and fat and then some recovery occurred the in 2010 as it’s the last year in there and a last period cavero first adjustment to the first improvement then occurred also sized raksha here we could see that in our sample one percent of the biggest firms produce almost forty percent of value-added and employ almost search is six percent of of labor which is in line in in general in general statistics for size distribution of Ukraine of firms in Ukraine and their economic weight also analyzing for dynamic dynamics we can see that till the crisis heroes off 2008 in 2009 it was observed Ned enter or firms and during this crisis year net exit started to emerge i calibrated parameters for this model in order to make it more precisely greynium case san clay enough they assumed that the rental price of capital is ten percent five percent for real interest rate five percent for depreciation but I analyzed real interest rate over the last buren decided to take median of this really interest rate and also I took more precise of a number for depreciation so I came to eleven percent rental price of capital speaking of modulus tissa T of substitution between plants variable where value edits I decided to stick to the value which signed clean off used speaking about elasticity of output with respect to capital wishes production function our production function parameters in in the original paper they share of capital was speculated as unity minus wage will share in the respond in the US industry and i decided to calculate these years on my own and i decided to use fixed effects a method to to to estimate value-added specification of production function now let’s move to empirical results so the first graph is distribution of physical physical productivity which is believed to be the x the actual the X the actual the actual tfp of manufacturing enterprises we can see that 0 here is them is the mean productivity so most into most enterprises underperform this mini productivity however only minor enterprises output from this this productivity it and it could be explained that those biggest enterprises which which which which there are few of them but they are the biggest and the most productive so they shift this average productivity this mean to the right and all and all the remaining firms remains to the left hand side and over the last period before I before crisis mean productivity shifted to the right so it’s it’s gradually improved and also distribution also improved so it became more normal but after crisis the situation worsened now let’s move to revenue productivity reparation vision this methodology is is the measure of of distortion so we could see that well that analyzing us do you let analyzing through the period which is under analysis so from 2002 to 2007 resource allocation slightly slightly improved by after financial

crisis as a variation of the total factor revenue productivity began to increase the this is a signal that allocation of resources began to worsen and also analysis or on some dummy variable show the small private enterprises are the least productive and the more distorted the most distorted industries are rated television and communication equipment industry refined petroleum products office accounting and computing machinery and the lead storage industries are food products and beverages rubber and plastics products and other non-metallic mineral products having having this these data this analysis in our I mean this analysis week we can obtain gains of resource missile occasion elimination so if we eliminate all the wages so in an ideal case if you eliminate completely all the distortions we can improve manufacturing productivity by almost from 97 to 135 percent however if the imp imp we apply benchmark distribution of distortions which is a us as bikes and Clennam it was computed that for us these gains approximately forty seven percent so the gains shrunk to 34 from 34 to sixty percent which satisfies our initial hypothesis hypothesis was was made for these management distribution gains during the last period allocative efficiency improved slightly but if we separate these periods in two periods so to pre-crisis period allocate efficiency improved by almost nineteen percent and then it worsened by almost twelve percent aggregate manufacturing productivity increase by seventy percent of 6.2 percent annually and taken average annual increase of allocative efficiency improvement improvement we can conclude that better allocation of resources is responsible for almost nine percent of aggregate manufacturing tfp growth then moving moving further if we assume that we eliminate all the wages we can we can decide that all the firms produce their optimal amount of production so we can compute how firms currently produce less then they could produce in an ideal case so we compare actual distribution of output and effective distribution of output so the extra distribution is this one which is inches more narrow however effective distribution is are these two distributions which achieved it to the right so on average firms produce produce more and it’s more wider which means that if you almost produce more or less more or less equally speed thus and analysis of these studio these distributions actual distribution and the effective one allowed me to conclude that most of the crania and manufacturing two prizes underperform the optimal level of production and most of them under performed by more than two times the lesson surprised the more and I prefer the more the level of under performances and speak speaking in terms of numbers almost eighty seven eighty seven percent of film firms and it performs the efficient level of production if we look through this effective efficient versus sexual distributions of production we can we can see that here are the quartiles x value added and here are the quartiles x in x employment at first to first two columns shows those theorems which underperforms and certain force force columns show those firms which / perform the effective level of production so looking through these tables allowed me to conclude that small small firms almost all of small firms underperform there they’re efficient level of production of IRS for big enterprise the situation is better as some of them could even underperform their effective level of production so the the highest burden of these of these distortions is

imposed on small and minion medium enterprises I also try to decompose to this this tfp are variants so the measure of of resource misallocation so at first I decomposed it by output output distortion and capital distortion and the results showed me that capital distortions has a positive and more or less stable influence on on resource misallocation whereas output distortions then they mainly determined didn’t the dynamics of the level of distortion so output distortions are more variable and they shift shift this situation so the influence the situation the most then I decomposed a tfp our variance by components so at first I become posted by between and within group components between group components means the means relocation between different different quantiles of productive BTW and within component or allocations of resources within firms of the same control of productivity so and this decomposition allowed me to conclude that the major relocations occur between within industry between the firm’s of difference of different productivity level and the major allocations occurred between the list and the most production the most production which means that while allocation of resources improves so resources from the list and the most productions productive they moved to the mean of distributions that us these tales decrease they did a bit between socha also I performs our robustness check used both here I used real data i also used nominal data results using change significantly also the initial panel was unbalanced i perform the same analysis for a balanced panel and results where and these potential gains were all ova as those forms which remains during the whole period of time so they does not enter does not enter they are the most lay the most productive so for them these gains are expected to build over so exit and and enter of firms they affect they affect miss allocation in industry also as I was as I was mentioned that we take elasticity of substitution between different to a mmm value-added equal to 3x and clean up from these arrow button checks by putting Sigma in equal to five and that’s unfortunately unfortunately that the property of the model that results are are increasing in Sigma so if you increase the Sigma results increase but but I decided to stick to this stigma which they took in their baseline baseline case in order to to make results more comparable finally working on paper I I realized so from very beginning that fixed effects estimation could be not very good for for estimation of production functions production function elasticity and working on this paper on updating I used Levinson between technique so just to recall basic results was abroad from 97 to 135 11 Petrine generated approximately ten percent higher gains so from 100 to 72 153 so probably when sampling is better and in updated version I will use this technique so as for now I have this this result also I perform some additional calculations so that’s the final part for exit enter so I I’ve got that those firms which which a less productive they are more exposed to to exit for exit those forms which have the hell higher he’ll hire the fpr they also have higher probability of

exit which which good signal that SDF PR includes these distortions so probably those firms which face mo distortions they have high probability of exit and those who enters our characters are almost fifteen percent more productive and has the fpr by such a six percent higher which we can conclude that this means that while the entrance the market so they have no have no ability to impose lower price so at first they have to import higher price and also those who who access they are almost twice less less productive than increments also I looked at market concentration so increase of human hair Fidel index by one standard deviation Lutz leads to slight increase of variance of TF PR by point seven percent so if monopoly power in industry increased it could lead to two more distortions in the industry also finally that’s for exports and imports so those those industry where is the share of exporters is higher they face hai hai district higher variance of the fpr however those were in a share of importers is higher they face lower GPR so they are more efficient and exporters and importers as as firms they are more productive and have less the FBR which means that they have enough enough ability conclusions enough ability to impose lower prices a domestic market so finally conclusions this paper makes investigation on the impact of resource allocation on a manufacturing productivity in a grain during 2002 2010 i have i’ve applied exeunt glenn of methodology and obtained results that full liberalisation is expected to bring from 97 to 135 percent of productivity gains whereas the u.s. distribution of distortions could generate us from 34 to sixty percent of of productivity gains also we found that more sir more than eighty four percent of firms and they perform their efficient level of production and most of them and we perform it by more than two times and finally improve improvement of business climate so if we rule out these these distortions they could be crucial for return to economic growth and as we found that the smallest forms are the most the most underperformance so probably it’s miss could contribute much to these growth if we eliminate these distortions thank you okay we had the change for the moderator on the go I’m gradually direct lecture at the sessions of forgiveness comments thank you valla dinner okay good afternoon already so I thank you Nicola for nice presentation and of course conference organizer to bring me here and being able to talk to many of you and to see my friends and to make new friends and to learn a lot about the reforms in Ukraine and not only as an academic from the data but also from the practical perspective and I’m going to discuss this very nice paper and I’m very glad that I’m discussing the paper of the one of the members of Institute for Economic Research impose a consulting who where I worked in early 2000s so I see that this Institute grew a lot and well has a very capable employee so I wish all the best to the organization so the goals of these papers are stated as to measure the potential gains from elimination of a particular kind of distortions in

manufacturing sector of Ukraine using a sort of very reputable metric of these distortions introduced by shame plan on ug 2009 but you should not understand you your work right so so you also extend it in a variety of way in particular you try to explain what’s driving those distortions by using for instance this technique to decompose hqk gains into some components which we can say well where they’re coming from right so so so some of these components are responsible so what’s going on within the beam of productivity within the quartile of productivity among the least productive companies or among the most productive but we can also look at the component explaining how this different group kind of compared to one each other right so so kind groups and we can dig deeper into it and see how it comes about and this helps to interpret this measure in terms of what drives these results and even next mccollough tries to relate those components in those gains to some measurable factors right so so and he names it to provide a policy toolkit for full liberalisation for reform in manufacturing sector this is most interesting and most relevant I think for the polity policy makers for academics to see how to we can advance the development of Ukraine and well I think this is very important and relevant study and let me great minds think alike right so so Nicola put the Doing Business Report ranking in terms of distance to frontier in terms of ease of doing business in Ukraine so I decided to visualize it even in gallah right so I put and those well all of us know geography of this part of the world right so so we know where Ukraine is and it’s here so the darker color means the darker yellowish and orange color means add the bluish gluer it gets it’s better right so blue means you’re very close to the best performance ever in terms of this metric the reddish means the worst performance compared to the best practices so this is the end of your sample 2010 so Ukraine was firing worse than always the lyrics except for Benny I think right so and then by this year ranks all of the countries in the region advanced reading the blue territory except for handful of Balkan states and Ukraine right so so there is a lagging behind in this doing business performance so Nicole asks what’s going on right so and basically he is very carefully studies the misallocation of resources how firms use and misuse maybe these resources in particular labour and capital in their production process and he extends the standard framework by adjusting the different work for Ukrainian parameters the cost of capital shares of production he popularized this nice work of the Institute to put together this data set of Ukrainian enterprises cleaning it and well what we saw that a lot of us a lot of people in this program really know a lot deep knowledge about the Ukrainian data so this is very nice and he offers a lot of stylized facts he starts a lot of style as facts about distribution of Ukrainian manufacturing and it’s well positioned in the literature right so it’s almost all work is done on late in America there is much less in Europe’s of my colleagues Sebnem and goralina they have work in progress about southern Europe using similar methodology and well most importantly and most interesting these findings drive some policy implications so let me elaborate what he finds right so if Ukraine moves to the u.s. efficiency in terms of dfb it might improve by up to sixty percent right but what is interesting also there is this nuance

u-shaped pattern over time where we see there are gains in efficiency over the early 2000s but then there is a deep right going back almost to the level of 2004 in efficiency because and Nicole explains that it’s must have been associated with the Cronus’s right and then so what is responsible so he finds that output distortions is opposed to capital distortion seem to be mostly to blame for these inefficiencies right and the most action the most reallocation of resources happening between groups of companies not within the group right so so it’s happening between groups of companies in the top 25 and bought on 25 productivity bins productivity groups right so not what is happening within each of the beans in productivity mostly driving these results so this is very a lot of kind of bigger extension compared to hay in a clan also so you need to push it more I would say right so so let’s discuss what can be improved to make this as far as the suit your thesis into a nice paper right so so and I should say so so I’ve been communicating with with with him and he probably implemented many of these things and when may be partially I misunderstood so so so let me tell so so I would say we should focus more on less technicality more practical advice rights also so it’s understanding ok so the academics are in this world you know in our ivory tower so almost right so so we operate in terms of you know terms like elasticity right so or I don’t know parameters you know heteroscedasticity but well if you want to be impactful we need to translate sometimes this jargon into you know practical advice what should be done so I think we need to guide so what measurable factors of the constitute this output distortions or capital distortions you provide a lot of editors but I think you should put these decompositions more to the appendix and put what is in appendix back into the paper I think that’s what’s interesting that is was different in your research versus you know so I would say you need to put more prominent role to correlations of these distortions with some measurable stuff rights also size ownership expert status industry so what you already do but kinda you know as almost sidelining it right so in police is discussed in section 5 right referring to very important work in transition and Ukraine should be measured put into regressions correlations graphs and into the paper I think so also to me as international economist is interesting so how did crisis 2008 affect enterprises was it external shock was at home grown right so how you know integration of Ukraine via trade links financial links propagate into this so so was it as I said external policy trade channel and well interpret those components move so also we been discussing at this conference who is the right benchmark so whom whose experience should ukraine emulate right so and I’ve been hearing and I agree that maybe Poland is bed benchmark maybe EU average is a better benchmark all right so so of course you need to replicate the result to validate them versus the US but maybe you can take the input-output tables of Poland which you know some there is some input output tables for lots of countries are available and do it now wage bill so so we can argue about this right so for instance even shake Leno they adjust Chinese wage bill for non wage payments all right so there are ways to do it in Ukraine I think it’s very important there are a lot of informal sector right so so I think well you probably need to adjust wage bill for informal payments and there are weights you know we know some proxies of shadow economy estimating factor shares by this regression it’s sort of subject to walk up their caves razor to critique so you do Levinson Petry in this paper criticizes even Levinson Petrine offering wooldridge extension of that so so it’s about simultaneous determination

for any inputs in productivity so you show it matters also you could improve weather even better this is the most important we need to see it and verify how do you calculate the table mark so I suspect you interpreted other way around so so so because HQ define the curve counts on the failure table 5 as the officials / actual size so 0 to 50 means actual size is too large so they should shrink you say they should increase okay so if you did it right this means that your result is in odds with all the findings for India China and with the previous paper who show that there are more small enterprises but they employ smaller number of workers at the time when efficiency improved okay so but we will see it and maybe I’m mistaken ok so now so almost finish so I also not sure what do you blow up load here on this graph is it music if it’s this or that so it depends but well just for your reference this light rights also so different policy implications will result from what we know about the covariance of this distortion so are they positive or negative I’m not sure from reading the version I read so unpossible robustness checks very quickly you can keep it so so so it doesn’t make sense to have the nominal data to me deflating capital stock by price of investment is could be easy as sort of usually people do right so so this could be done and then I’m also very interested so what is going on for average firms within their so is their firms are firms moving from distributions into the middle or they always jump between extremes of distribution so so and what happens to market shares it’s a very careful implementation of original HP a paper for Ukraine with nice extension lots of nice exemption extremely important for Policy more stylized facts and economics instead of technical stuff please thank you let’s probably collect some more questions and then we’ll give a couple of unis for my culture response yes please yeah all I have religion this is super paper frankly and if in the past 25 years that I’ve had experienced in the crane economic reforms in policy had moved as far as the study and research and economics at this paper it’s like we wouldn’t have a problem today let me make two points and sort of a in question form well the first was just a comment I there is a huge amount of policy implication there and it’s very timely because today like in many developing countries on the verge of maybe moving forward there is this common fear almost universal that we don’t have any comparative advantage in the world which of course is logical nonsense this paper goes a long way to say that there’s a huge productivity potential to make Ukraine a big exporter on the substance of it I wonder whether I miss something or whether there is another element that you have missed you worry a little bit that you are overestimating the gains because not all distortions will remove short but you may also be under estimating them because I’m not sure I see anything in your methodology in your model that really captures X efficiency gains certainly the literature of the past 50 years on exposed estimates of how much liberalisation gave first of all it has always been on partial liberalisation almost nobody had two full liberalisation often shows that the actual gains are far greater then then early estimates would have indicated so I think you needn’t worry as much about that in at least comment maybe you got it there maybe it’s wrong I think this is best summarized by something that a well-known personage in international trade theory once said max Gordon and he was in Poland together with their with me on a World Bank mission in explaining

to the minister of industry who said well we have no comparative advantage how can you tell us that things are going to improve a little ization well max Gordon said you know what be prepared for pleasant surprises thank you very much do we have you can keep it do have anybody else who might ask some questions us know because i have several as well I frankly speaking I have been working with this kind of data since 2003 and some of my students and some of other students of the program also have been working with his data and Nicola probably the first one who continues working with the data after gradiation so I cannot express you how glad I am that there is somebody else in Ukraine who also works with the firm level data at this very very high level because academically this is extremely highly a very high paper I had a couple of pitches over here first this some technicalities I couldn’t understand how you treated outliers it looks to me that you started estimating that your production function actually before you treat it out loud before you tweet it out fire so it looks like if you a since in Ukraine we have home aspirated distribution of employment allowed potent of most of most variables those outliers can affect both the distributions and both estimates of the distributions distribute estimates of the particular coefficient so I thought you might consider treating the outliers and kicking out of them them out of the regression before you actually start estimating TV because it sounds like one percent of the firm’s seemed completely different from everything else or maybe two percent but you might might you might look at this as well another thing you you were talking about coke and petroleum industry as one of the most distorted and I think this is note both to you and to myself as well this industries has one of the smallest number of the firms and that industry is one of the most diverse in terms of the average size of the firm understand the dispersion of the size of the firm so probably this could drive your results because if you if you’re looking at the food industry the food industry handsome 55 thousand firms over or per year and this coke and petroleum has less than 100 transfer of the firm’s per year so this could be the reason why those results look so much different so once you interpret the results probably the size of the market should be treated somehow and at the same say and the same argument goes to for example them with the monopolistic power of some particular industries and concentration of some particular industries because in some of the industry’s concentration is huge and this huge concentration basically means that you have some company with a huge monopoly power which looks like it has huge Mahna huge efficiency gain where is all others like very far behind but this is not because they are not efficient but this is because they do not have a dis market share and I think this is once we are talking about efficiency efficiency should be somehow treated separately from the market shares so this was my kind of most ooh those two questions that i wanted to show do you have any other questions yeah yeah sure practical suggestion i agree with Vadim that there’s there’s a lot of policy content and it should be you should market this but I wonder whether a better way of doing it is to do write two papers any paper that you write for policymakers that has even an appendix with equations maybe quickly get put on the shelf anyway so just just do a policy paper for heaven sakes okay you have any couple famous to respond to have 23 million some since you have a mic so thank you very much thank you very much for listening for commenting for asking questions so at first I would like to to comment a presentation by the team so I agreed to to both his process of my of my papers on and with all of my with all of course of my papers as as these map it seems that we are thinking in one way so to tell the truth I have the same ideas on how to improve this paper so here I have I have no I I have nothing to add actually nothing to it so I agree with everything so speaking about the

comments on on other different on other sources of these gains so I used just the basic setup of C and glenna model so i decided to use these the baseline model to perform this this exercise however there is the recess law of literature where some where some other some other distortions are included sample its implications are included so i’m aware of this literature but this paper i decided to start from these basic from these basic model as here here i used to factor model there are already extensions with sri factor model is still working paper but but it was met by Bank of Portugal so I presented the results here also so let’s is under under under my research so I’m doing it so I’m extending this methodology and answer in Volvo dimmers concerns so yes you’re right so I have evaluated these alphas at first as it was suggested by sinclair know that we should we should estimate alphas calculate productivities calculate wages and then based on these productivities and wages dream outlier so we dream outliers not by size not not by not by input but by productivity and wages and to tell the truth afterwards I I didn’t recoleta to alpha so probably I should do it once more second iteration so I also agree with you concerns EDS time is out that’s all okay thank you very much I declare the session closed thank you