 Three questions provide You know let the panel respond and then and then we'll take another set of three So let's see Okay, so let's see let's let's start we have okay, okay, we'll go start here Yes, we need to get a microphone. I thank you. Yeah, there are logistics. Thank you very much My name is Matthew Powell from policy management. Um, I like principally well about principally about organs and So I was very pleased You were 1993 GDP data was a reliable point I was there and I I Could talk about this for hours, but I won't You're well the point the What's actually I think that anybody who'd done what you have done would have come up with similar conclusions and That's there are a whole host of reasons for this And there are some technical reasons I think the political economy reason Justin was talking Probably all more important because the technical solution won't work unless the economy is But when you're trying to analyze this further and look at the Reason what I would say is what to look at who's who really is using BP Dana When we Compared GDP Hey, you know that the Monetary Policy Committee will use it to make a particular decision Every quarter that quarterly G the accuracy of that quarterly G very important and That the whole of the the rest of the system how it's designed sort of hangs off that and there are other users That one key user Who is the key user in the country? Will you've answered that question? Second question Let's see. Let's go here Okay. Hello. My name is Katrin Rusner. I'm working for the German Development Corporation and I have two Questions or remarks the first one is going to the first speaker who called us kind of part of the problem and I was wondering a very simple question. Did you get any positive reactions on your book from? Developing countries or African country represent representatives The background to this question is we've been working in Central Africa with tax administrations and I thought they would be really happy to have even that national level the data on Specific sectors or even GDP reliable data And then the second comment as to the third speaker Just a very detailed thing When you try to associate Participation and collapse and associations with GDP. Have you thought about explanation that often? associations or NGOs are kind of Self-employment measure. So that's maybe why it's not correlated with GDP One more question. Let's do Martin Thanks. I have questions for both Morton and Justin Morton, I think it's really important to do what you're doing, but it's a little bit frustrating I I hope you also present this to Policymakers finance ministers finance ministries paint people making decisions about the funding of statistics officers And I'm not awfully surprised that you're getting a back Backlash from stats officers, but one question is why Why do we see these these for example long lags and implementing SNAA or they vary between countries the rebasing which are pointed to and one obvious question would be well Is it related to public resources for the stats office which should be public data? Is it the case? Is it just a matter of the finance ministry allocating more money to the stats office? Or is it something about what the stats office does? They will be a terrific question if you if you've addressed that to tell us a bit about it and another suggestion is You know, I think in this room we mostly got micro empirical economists So it essentially was preaching to the converted that we were already either know this or think oh, I'm no surprise at all but Macro economists growth in Perix You got a whole literature a thousand papers probably running growth regressions on cross-country data I'd love to know whether there's correlate whether the correlations between these these lags and so on these quality measures and The right-hand side variables of those growth regressions And what sort of to learn about the structure of measurement error and the potential biases in this huge literature? Being a micro person, I'm probably inclined to believe there are such biases I'd love to have some more evidence on it on Justin one quick comment The if you be very careful about arguing that that cost of basic needs poverty lines are priced indices Because you've got to remember that there's a non-food component in the CBN poverty line as it's typically calculated And that non-food component is calibrated to an angle curve and that angle curve is shifting over time for all kinds of reasons Changes in tastes changes in activity levels and so on with implication being that Maybe what you're seeing is that the adjustment using this the CBN poverty lines is Actually overcompensating the real poverty line is rising over time because of the shifts the angle goes I'm not saying that's that's the case in Tanzania, but that there would certainly be a possibility you'd want to Be aware of thank you. Let's start with Morton and then move on to Justin and then we'll do another round of questions No, it's all thank you That's a quite a few questions, and thank you there are very good questions, too Let me first think start by the simplest one. Are there any positive reactions? Yes, plenty the kind of reactions I've gotten and the political and public ones are the ones which are coming from the directors of statistics It's a quite different kind of story the debates and the kind of encouragement I'm get privately and publicly from those who actually are involved in doing national accounts Which do feel one of things which implies some to the question you asked Martin that they have been on under prioritized in many ways So so there's a lot of a positive thing here as well Some are public some are private Moreover, there are many Wested interest in terms of African development Bank would very much like to see investment in business registers Etc an updating of those across the region They're happy about the news, but they also have to play the political game of not Of trying to be on the same the team of the right people You neck I did invite me because they wanted an investment in national accounts as well So there is a some positive reaction so far now handling this I hope one can do a free debate on it There is two kinds of Things you have to think about at the same time one is thinking about the National Statistical Office as an institution Which has been if you think about it Been in relative neglect as an institution compare across the road to the central bank for instance Which has been secured legal independence Funding independence and has a high-rise buildings with computers any kind of survey could could could show you this the National Statistical Office By comparison has been relatively on the funded and hit less and it's kind of for those who are interested in growth empirics It's kind of puzzling irrespectively that IMF and the World Bank invested so much in Reforms that were supposed to induce growth. Well, I'm not at the same. Well at the same time forgetting to put in a baseline Estimation that we could anyhow reasonably think that we could could measure this Then that's on the National Statistical Office, but there has also been a shift of funding and this has been Shown by many other studies as well where there is a migration from economic statistics to social statistics Now this is has to do with other types of local incentives at the local ground, right? So that means that there is a lot of our data some of it micro some of it macro on social data and so forth It has to be Survey funded so there is a classic mistake in the Millennium Development Goals agenda where we have the eight goals the 18 Indicators and the 48 targets we kind of forgot that this is might seem to us as administrative data data that are available to governments on The day-to-day operations. It's not it's survey data for these poor countries We're looking into so that means that you need expensive data exercises that means Funding is coming into National Statistical Offices that is coming in very particular form It's coming ad hoc and it's often coming as I per diem reward for data collection Which is not what national accounts do which is not what economics the statistics do they? Collect data within the administrative units have have and and our as any national account specialist would tell you is that you know A omnivore of different types of that statistics Unfortunately at the moment most national statistical offices are working more and more as a de facto Data collection agency for higher than one is actually providing those kind of data that may be not microeconomics needs But states would like not only do we have a knowledge problem among the macroeconomists We do have a knowledge problem at the central banks for instance to address that question directly one It's a frustration at Malawi and Tanzania in Nigeria expressed about this that they cannot trust as their own data coming out Of the same place because they know that the statistics are Not good enough to base their the decision. That's again related to that kind of shift in funding I'm not saying that it's more important to have GDP statistics than in font mentality. I'm not making a Normative statement about what we should be counting on the counting. I'm saying there will be a demand for data the data Supplied what matters is how that's data is supplied and by why methods and how it also then Distorts the the type of Facts we get so then finally then that means that we also need to think about Very very carefully about for instance things like we should only pay for results. That's a very very naive type of of It does sound well Technocratically, but it's extremely naive when you know actually how fragile these statistical systems are which also the work of Just in I think shows but I'll stop there and I hope I touched One one thing I forgot endogeneity. So yes, there is a what you said There is a of course a correlation between Income and base year which you could which you could tease out There is also more over a when you see a GDP on the time series. This is more apparent You are seeing like a state so it means that when you are in Tanzania for instance The informal economy is illegal. It is also not covered by formal accounts when it becomes legal It becomes included in the new informer sector council. There's a lot of political economy There's a lot of endogeneity and one of the central messages in my book is that these are not facts objective facts But they need to be understand that's con us products And therefore we need to understand the political economy in which they are embedded in order to use them and also understand How to produce them better in the future. Thank you I mean a quick response to Martin Thanks, I think of course you're right. I don't think the cost of basic needs poverty line is a perfect CPI substitute And I'm trying to think through on my feet about if what's happening in Tanzania is food prices are rising rapidly I don't know a food as a whole is a given good I don't know what's happening to the the proportion spent on food, but we'd have to look into that But while the CBN is an imperfect substitute for the CPI I mean in Tanzania's case the CPI Apart from its theoretical disadvantages of not covering rural areas. I mean it's just clearly Wrong and through the period of rapid food price inflation in 2007 2008 9 was was capped at 10% as a rule And since that CPI series has been revised and and shifted so I don't think we have a perfect alternative But I think the cost of basic needs poverty lines provide an interesting Independent measure of what's happening to to prices But I would be I would be curious to get more advice for you from you offline about what to do about the angle curve situation Let's go quickly. Let's do a fin and then the woman behind And it's nice enough to let me go first Organizer at the risk of making enemies. I want to make two points. Um first first There's an extraordinary extraordinary amount of cross-country heterogeneity within Africa and one of the things about Ghana and Nigeria that I find so weird is that Especially Ghana they they supposedly have so much money speaking to this issue of public resources But they have some of the worst data in Africa And it's not and and and you take a country like Ethiopia one of the poorest countries in Africa And they have one of the best well-run statistical agencies in Africa So so there's a big puzzle there and the second thing has to do with macro versus micro data in in Ghana and in Nigeria the macro data and the micro data are equally as bad and In defense of some of the national accounts data In a previous session where I showed some of my results using the DHS data people seem to think is is Good quality. I found very high correlations between the employment shares using national accounts data and employment shares using the DHS data, so I Think I mean I do agree that there's a need for better quality data But I think we need to be a little bit more careful about saying Africa has bad data Thank you. I Kind of have a couple of observations. I mean But let me first make a small advertisement Jeff round has actually four wider written a very insightful Working paper on the situation of statistical offices in Africa. Those of you who really want a Balanced insight from somebody who really has spent her it's old career profession Doing this you might want to take a look at that Now the next point I want to make is that weaknesses of statistical offices is not just something that affects macro national accounts data. I Mean that affects across the board It's not just one part of the data collections that we are doing that's affected Now the second sorry that was a third thing which sort of kind of posses me. I mean The title of the book is African growth miracle or statistical tragedy Yeah, sure number the paper but but but the point is this so are we supposed to think that there was no growth? Yeah, I have but but the basic point is just that the situation is that We are all trying to work with what is available at this moment. That doesn't mean that we take it as Objective fact that means that we approach it as economic analysis as analysts who try to look at these data from different perspectives if you take the growth and Poverty project that we have been implementing over the last three years what we've been doing is We've been trying to bring different data sources different observations together and then try to tease out What's the story? So I mean, I'm you know, we have we are completely united Statistical offices in Africa need very substantial upgrade Very substantial and of course it would be great if Bill Gates would kind of start put some money into that At least in some places where there are lack of resources But but but that doesn't mean as I understand it that we can just basically sort of throw everything out We need to do the very careful Assessment and try then to sort of understand what we can use and what we cannot and cannot use and Maybe as a slight comment to Martin I mean what basically comes out of trying to do what what you just suggest it should be done is that yes There's a lot of sort of stuff that how can say It makes it problematic in the very short run when but when you look over the longer run Then the things actually tend to be quite correlated Thank you. Yeah I and and that's partly You're the host but I still want to disagree with what you were just saying and ask Ask it to do to come comment on the issue of accountability I don't think it is it is quite that that benign I'm thinking of has the international community come to when I think about DHS have they contributed to creating parallel institutions Where they have and and that in itself is is fine But what has that done to the mainstream institutions and have we so so my example my example is is 2000 PR speed Tanzania my colleague and if it was at all we need to do we need you know we need to support a household survey and I think the the census was was over ten years old and my instinctive reaction was well Shouldn't we be doing first the census that that's a debatable point But I was very unpopular when I asked it because the international community needed the household survey for for the PR SP Now that potentially then is actively undermining the build-up of a local system that is needed to make the right decisions and The research community that that analyzes the data is not is not the cause of the problem But contributes to that and and I would I would still say that it as research is still our Responsibility to look at in the end is we're interested in development And if we think that monetary committees are an important are an important Institution in a country that we need to work harder to make sure that those Institutions locally are supported and that the accountability is to the primary users I eat the people who put this the benefits from from monetary policy and not an international community I'm not saying anything out of the ordinary that simple Paris principles, but but but and that's in a way I want you to do in wait, what do you see more than what they get to ask, you know What do you see a way forward in which in which we can contribute to ensuring that there's an accountability in In in the building up to these students at a minimum that we don't where you fall it undermining them Let's let Martin and Justin. Do you have final comments in 30 seconds to a minute? Anything to do you want to finish off with or First of all, yes, it is extremely obvious that there is heterogeneity In GDP and quality of statistical offices and indeed that my I mean, I think that the tables I showed showed exactly that it goes beyond that that's just a surface kind of thing And I think also that the kind of discrepancies your or what for you you said are maybe surprises It's not so surprising from an economic history perspective if you think about one of the things we are thinking about there It's that lack of formal recording is also correlated more or less across the board with with poverty in Africa The correlation is a bit different because also there has been a historical way of building states in Africa Which has been less dependent on direct land titling States that have been less dependent on direct taxation as well Ethiopia is an outlier in this with with both of these things that it stands out Rwanda Burundi as well and there is also the southern worth the western Africa is also distinct in that sense hot taxes was widely collected in southern Africa where there is no hot taxes in Southern of Nigeria or in Ghana as you point out which matches up very much with long term Credibility and learning by institutions. So these patterns can be very well explained and it's also part of what I touch upon in the book second I'm not saying Finn at all not even close that we should throw everything out. I'm saying exactly the opposite I'm encouraging you to listen carefully to that so you don't misunderstand that I'm saying there is a knowledge problem I'm saying we cannot throw out the GDP. There is a need for these aggregates Not only for us data users session on the knowledge side There is a need for these aggregates on the policy level and I think we need to think about Exactly that issue of accountability and I thinking about local demand for data and global demand for data One of the things we have done and that is this problem started to be bigger and bigger in the 1970s When national statistical offices were not able to deliver what the World Bank and IMF was demanding of them and therefore often sidestepped the now investing in the national statistical offices and did other types of quite quick fixes Negotiated numbers or indeed collected alternative measures and I think we need to think about I think the way forward is to invest in local accountability I think that they brought this is not only about GDP statistics But the ability of the state to know something about themselves is quite important the ability to conduct a population census to know whether the population is 130 or 160 million does really matter for a lot of other policy variables and therefore we need to think about When we think about Demanding these data which we do we need to think about how to invest in those sectors the supply of the data I'm not suggesting to throw it all out as well and then I hope you don't misunderstand that I hope also statistics statistical offices across the region is is Suggesting that what my suggestion is that we need to pull them out of the relative neglect I've been for a couple of decades and think through this very carefully once again Thank you Well, thank you all very much for a really stimulating session. I thought that was extremely interesting We have the break. Oh, and then we have the poster session Please please go there'll be seven doing in a row. They'll talk for about 10 minutes each You can listen to those presentations or there will be speakers standing in front of their individual posters. Thank you