 been trying to go through the exercise and if they can show the results or if they have some difficulty before we enter into the session of today. Please go ahead Mr. Payudul. Yeah, yesterday I tried to calculate SDZ 2.3.1 but I faced some problems. I have tried to calculate using his data but what data are available in the dataset, what data are available and what are not first I discussed then what problem I faced I will discuss. In the questionnaires I have seen in section 7 part B production of crops and its self-price available against household within the last 12 months. It is available and like livestock and poultry and its sales price is available. It's okay but the data which are not available no data available on number of days work against the product and number of data no data available on production cost of agriculture product is that are not available. These are the problems to calculate the SDZ 2.1 but and another problem I faced in the dataset we collect data against household not for its producer in that case household will be the its producers. I don't understand this again and and number two problem faced constant selling price of the product that does not mean constant selling constant selling price we collect nominal price in that case what what do we do for the constant price second price the second problem is that and another question I arise why we do not consider the by-product of this product it's another it's my questions and another questions I feel how do you calculate household weight in is in your slide 14 that you have shown in the last day these are the questions I feel if I got this problem then if I got the data available then I think I will estimate the SDZ 2.3.1 and 2.2 2.3.2 then what is your solutions first solutions you you have a I first face some of these producers but we have collect data against household in that case household will be the small scale producers first this problem then how do we calculate the available number of days work against the product these are the problem what is your solution okay so shall we go through these questions before before you go to answer this you know questions you know let me add something on on top of this this highest is you know the target of highest household income expenditure survey so this has a sampling frame and it may miss some of the aspect of this you know identifying the uh small holders so we should we should have to think that one but as there are information that we can play you know incorporating those and calculating the indicator at the same time what I remember I was you know saying that you know plot wise information would be required you know one plot what was calculated what was the price something like that now in the household one you find aggregate the total area they have used yeah production they got and the price so will it harm this india no calculation the indicator this is another part if you answer collectively okay so shall we go through these questions because there are several questions here yeah please please before taking another person yeah I am not sure I got all the points but if I understand first of all the main source is the household income expenditure survey that you are trying to use the data coming from that one was uh depending on the survey the statistical unit could be a household or could be a holding now if you take the household survey the statistical unit is the household so what we are concerned here for this computation is the household which are involved in agriculture so for example we are talking about the lsms isa from the world bank and the component isa means exactly this is a component dealing with the agriculture aspect so we are concerned with the households engaged in agriculture because you can have households engaged in different things in the household income expenditure survey but here we are concerned with the households which are engaged in agriculture so we narrow down to this household and this is where we can make the computation because for the others you will not find the variables of course because they are not engaged in agriculture so so we consider household as a producer but if it is engaged in agriculture yeah yeah yes you are you are right we make this it's a kind of approximation you are right so we approximate so the theory is that in rural area almost 90 or 80 percent of whole dings are households are engaged in some kind of agriculture so we assume we approximate the the whole dings the households engaged in agriculture with the whole things it's an approximation yeah you are right on that so we narrow down on this and then we do the computation on this this sub-population of the household slash holdings that are engaged in agriculture agriculture in the broader sense so it's crop, livestock, fishery and agroforestry not the forestry in the open but agroforestry so another question is i have another question we have used the volume of productions of crops uh livestock but why we do not consider the byproducts of this? I think the byproducts are listed if you look at the list of oh we consider okay if you look at the list of the products is included oh we also consider the byproduct okay okay then it's okay yeah now about the prices i do not understand very well what you mean by nominal price and over prices oh i i say a farmer produce rice and it's and it's current price we collect current price but what does we mean at constant price okay now perhaps i can try to answer that yes so i think what is meant by constant price is related to the fact that when when you implement a survey you don't collect all the information at the same time but like for example the survey is implemented in a year time or in a six month time frame so the price reported by the by the farmer may depend on the in fryer let's say inflation or like the the price variations that you are doing in the year so uh normally like the price that is used is deflated for this infernal inflation so you take a sort of average price like a middle year price let's say or a price referring to to the the middle of the reference period now if if that's i'm sure that that is not done systematically in our country so if only nominal prices are available or current prices i think they can be used also because like uh in infernal inflation could be not very high in in most countries normally you have a very high variation when you have some sort of shocks big prices or so yeah i think this is the i think that's a good explanation is that okay for you yeah yeah now i understand so we will we consider the inflations of data okay and then we yeah if you have something yes to correct for the in infernal inflation then it would be it would be better yes yeah and you know when you are dealing with this work you work a lot with proxies because you know the ideal situation does not exist in many countries ideally you should have the the crop and the price immediately but most of the time you don't have that situation so you need to manipulate somehow the data to get as close as possible to this what you want to put what is in the formula but the reality of the formula and what you find in the ground are not are not the same so you try to get as close as possible with some statistical techniques in order to make a consistent calculations so the price is a difficult one and even if you don't have the price for some crops you go to the far start so this is even more remote than the country price so yeah you need to to to to be a bit flexible on some of these parameters okay any of the okay there was one question from Amirul about the household hi and the household income consumption survey based on household listen the idea I was mentioning about need for growth level data I mean I think that's not the main the main issue why I was saying the growth level data in most of the cases agricultural service starts collecting the data at the growth level you know then we are bringing it to the holding level as you have seen yesterday therefore as Naaman was explaining in most of the cases the holdings in the households you know about 80 percent are you know much therefore at the end of the day we are bringing at the holding level aggregating even though the data is collected as a plot therefore if the data the HIC data has you know an aggregated value at the household at the holding level then that will be the starting point to to start computing the indicator as long as we have all the required variables to compute this indicator through this survey okay any of a question in the Q&A any of a person try to to compute these indicators and what are the issues that you had and another question is here how do you calculate household weight say in the in your in the last presentations in slide 14 you you have seen household 1 weight for household 2 weight 5 household 3 weight 10 how do we calculate this way I would like to are you mean the sampling weight yeah yeah sampling oh that should come from your sample sample sample design and your sampling experts should provide these two very statisticians and and IT people so it depends on how you what kind of sampling design in your presentations last day in in in your slide number 40 you have seen how household number 1 against weight 4 and household number 2 it has weight 5 how do we calculate this yesterday you have seen in your PowerPoint presentations yeah that's what I was planning that the sampling weight is a reverse of the probability of selecting the unit now this depend on the sample design that you are applying in this survey for example as I show I indicated in some of my slide in many countries you use two-stage sampling these are these these weight is okay here is a household weight against a household every household has a another different yeah different weight how how do we calculate this different way is easy let me let me you know clarify you know mr. yeah you know these are the small example so you found that in a household 1 has one weight household 2 has another weight but in your big samples that are said within the PSU all of the household will have similar weights yes and the second the second PSU will have will choose a similar but so oh the exact weight that is a that is the individual within the survey it is you will use that one okay it is not household weight just an example okay okay I think say it is a household weight then yeah yeah this was just you know these these two households are coming from different sampling units yeah that's why you see different weights as a household label otherwise as I mean we'll explain it very well all household this is the same sampling unit will have the same weight yes so I have potions in his data we have 16 startups so we can find you have 16 different weight but here he said then I have questions no no it depends on your design so yeah so the strata usually you know you have two stages and you go below the strata so you have a I don't know if you use the animation areas for example and so you stratify your universe and then I have seen in some of your surveys you distinguish between the livestock domain the fishery domain etc etc so to me this looks like a stratification I would like to know about if a household has a have more uh more livestock and and another household is a low uh livestock in that case weight will be different now in that case weight in the same stratum by in the same piece you then it will be faster you know we are not we have no right to handle you know or modify the you know weights the weights that comes with the sampling that should be used as it is so it does not depend on it does not depend on how many it is related with the selection probability and the reverse of the selection probability okay so that will be that will be as it is in the coming in the survey okay I think that's a good explanation and anyway I hope that after this training we'll have a chance to work together and to to look closely at the data and see how all these problems you can discuss in more detail that after this training probably we can start looking at the data the real data and see how we can compute the indicators so if that's all for the questions shall we move to the so yesterday just a brief summary yesterday we focused on the computation of the STG 231 and 232 the methodology and particularly the practical exercise with the steps shown by max and then the illustration with a demonstration numerical demonstration by Audrey so today we will deal with the another difficult part of the the issue is the data issue so to introduce the session of today I think either are you ready so either I will make the first presentation this will be followed by I think importantly presentations from bbs two presentations from bbs talking about the surveys and data gap and then we can see how we can discuss and find solutions for this data gaps either you have the floor thank you so before before that one I know for it can you just you know include mr aloudin and mr kamrul islam as the analyst so that they can they can you know no problems can they can they rise their hands thank you aloudin why and mr kamrul please raise your hand so that they know they can identify you aloudin alzad I found him yeah okay I don't know but some technical problems and mr kamrul also raised your hand so that I don't know let me know yeah please go ahead thank you so um let's say that with day one and day two we close all the main let's say technical part of of this training so we saw how to identify small elders how to compute the two indicators and now with with this last presentation we focus on the main data sources although we we have largely talked about that during these two first days and on data desegregation required for these indicators so I will first provide a recap on on the important data items needed both for the identification of small kid food producer and the computational indicators I will try to be brief on that because you and you have heard enough on that and you will also have our slides to to go back to what has been presented then I will discuss the main data sources and some FAO and other international initiatives that that are important for these data sources and then I will discuss the the data desegregation in general let's say in the context of the SDG monitoring framework and specifically for indicator 231 and 232 in doing so I will briefly present some guidelines that the the office of the chief statistician of FAO has produced to to support countries to produce these aggregated data and I will also give an overview of the type of technical assistance that we could offer if you are interested to for example to produce these aggregated data of indicator 231 and 232 but also on other like FAO relevant test each indicators so I think so during the the first two days we need the like the data items that we need are needed to identify small orders and compute the two indicators so to identify small orders you need the data item on the over the land that as we saw is given by the sum of cultivated land with permanent crops cultivated land with temporary crops and fallow land and and it exclude all the the other type of of agricultural land such as for example the the land for buildings the forest land these are excluded from the competition then for livestock you need the number of animal needs stock and conversion factors to get the TL use for the data on revenues you need the data to compute the value of crop production livestock production and forestry and fishery production and the the revenues components have been already presented in day one but also in day two but I included them here like as a recap so you can also refer to this powerpoint presentation to see the components of the crop revenues we have already discussed this we needed the livestock revenues and fishery and forestry revenues then the data item needed to compute the indicators so for what concern 231 and 232 231 we have two components the numerator of the indicator and the denominator so the numerator of the indicator is the revenues and the denominator is the labor input so let's see a bit more on how like which are the data that we need to compute the agricultural output so the value of the agricultural output and the data that we need for to compute the labor input so as we said before it wouldn't make any sense to quantify the the agricultural output in terms of quantity produced because we have different types of products that are produced and and we cannot send them together like I don't know apple with with oranges so we have to look at the monetary value of production so for doing that we need the let's say a quantity multiplied by a price so what what we have discussed before and the monetary variables need to be deflated so deflated this what what we were seeing before in theory you shouldn't use nominal or current prices but you should use the deflated prices taking into account let's say the mid the mid-time of the the reference period of the survey and then this monetary value should be standardized using PPP conversion factors that you can download from a website we have put the link on on the previous slides that have been which you put on the slides for day one and slides from day two for what we're trying to do is there are various ways to measure the labor input like for example the number of workers you use the number of days for the number of hours worked of course it would be much more precise to measure labor productivity in terms of like a unit of of labor so the in terms of hours worked the problem is that very often in surveys they are not available at this level of detail so the solution that has been adopted is to use the the number of days worked so you will have like a productivity where the the unit is the the labor day let's say so as you know in agriculture you have like situations where people work only part-time in the agricultural sector and they also have other occupation so ideally we should be able to compute the number of working days in full-time equivalent so like the working days that are equivalent to eight hours of work so to do so you would also need the the number of hours worked by like an average in a typical working day on the old thing if this is not available you can simply use the number of days work and all the type of of labor are considered so they paid and paid labor including family hired or exchange labor so we don't like really matter if the household or the farm is paying for that that that labor but just to see the amount of labor that is used to get to a certain agricultural output so now we have indicator 232 that measure the average income of small-scale producers this indicator refers to gross income so an income that is measured through revenues minus cost plus the stock variation when when it's available this computation like for for problems of data availability and the methodology exclude the depreciation of asset taxes paid by by households of farmers and also the employment-related obligations such as for example I don't know like pension paid to employees or this kind of things so the revenues shouldn't include only the quantity sold for the quantity that is produced and then sold on the market but also should include the quantity that is consumed by the farmer and the quantity that is stored for a precise let's say review of what should be included we we have another slide and then also the cost like the cost should include everything that is used to help let's say around the production process so normally it comprises all the types of expenditure that the the farming core in while while producing so again in this presentation I put a recap on the main components of revenues and the main components of cost I think now one has already like presented these slides and also the next one on other sources of income so we don't need to see them in detail what you will will notice and and that we look here is that you have some components that are put in the revenue side and in the the cost side this is because some like for example let's find let's find an example okay so the crop use for paying rent why it's among the revenues because if I don't use for example the my production to pay for the rent I would need to use my my personal let's say income but then so I this is part of the revenues but this is also a cost because it's like I need to rent the land in order to produce crops so in the end these components are somehow like they are eliminated from the final income competition because they are counted both as a revenue and and and as a cost so you can decide whether to put them in equation or just remove them from from both sides and this is this applies also for livestock income and fishery and forestry income so which are the main data sources to compute this indicator I think this should also address some of the questions we received on the exercise from yesterday of course like each country should use data that are available in some countries for example we may well developed agricultural survey that include information also for not only on crop and lives of production but also on the amount of labor the labor input use the cost of production the all the all the data that are needed to to compute the indicator so in this case the the unit of the analysis so the unit to which indicators would be referred would be farms that can be bought in the agricultural areas are in the household and non-household sector okay and an example of of good agricultural survey that could allow to compute this indicator is the agri-survey our service design under the agri-survey project I will give more detail on this iacobas also talked about this and also naman and and perhaps you have also already engaged with the FAO team in charge of this of this project then when instead an household survey was as in the case of Bangladesh I think because for example you see this rural library information system project is a project within the FAO that try to use existing data to compute some of these ethnic indicators and and within this project they have used the data of the of your household income and expenditure survey to to compute as each indicator 232 so perhaps we could see also how this data have been used and and and if this survey could be let's say complemented to compute also as each indicator 231 so in some cases a good data source could be an household survey integrated with a module on agricultural activities of households in these cases that the statistical unit would not be the farm anymore like in the end it would be the farm but the the initial statistical units would be the household and then among these households all the households with agricultural activities would be considered of course the agricultural census could also be our source only if like all the needed variables are included in the census and in some cases also administrative data sources could be used in integration with other data sources although this let's say is not the preferred source as it's it is very rare that administrative records will contain all the information that we need for the computation of such indicators and here my colleagues can correct me if I'm wrong but I think it would be like more likely to have this kind of data for commercial farms than for household let's say than all things in the in the household sector like administrative data for for this type of holding so these are the four main possible data sources as a rule of thumb so what we were saying in the previous day and this is perhaps the problem that we could have in Bangladesh so all the information required for the the computation of these indicators and the specifications of small older should come from a single data source to yeah to avoid confusion between production units and especially like to be able to trace back all the variables to a specific household because we are not looking at average measure at country level but at household or farm level so it's important to have all the variables referred to the same statistically of course there are and I will briefly talk about this there are statistical methods that allow you to integrate data from different data sources but the fact is that the integration of data from different sources should be ideally planned at the design stage of of this survey so like normally when you want to use multiple surveys for the computation of a specific indicators this should be planned when the sampling design of this service is planned when so so here a solution could be that when all the data required to compute in the indicators don't come from a single survey then we can redesign a survey that collect the required information so let's see what could be could be a solution here because the idea is not that of having a completely new survey but adapt existing survey to to collect the required information and that's what the agri-serving project normally does so by by the realization that in many countries the agricultural surveys are seldom conducted in a systematic way but also by the realization that this survey often do not collect all the data that are needed for policy making the agri-survey the default developed the so-called agri-survey model that I will briefly describe now and and normally when this model is implemented the approach use is not to let's say design a completely new survey but is to go in a country see what data are available and what surveys are conducted and see how these surveys could be improved using the agri-survey model and the idea is that of basically bridging the data gap sorry the tenure gap that normally exists between one census and the other to have more regular production of agricultural statistics the agri-survey is structured around the core model so there's a core agricultural production model that should be implemented every year and this core model is then integrated with some what we call rotational models so some models that are not administered every year but on a rotational basis and these models cover different let's say domains there are there's a labor model and economic models that collect for example information on cost of production there is a model on like the equipment so the machinery is available for the farm and so on so on a rotational basis a series of information are collected and and this this survey is designed to collect information on indicator two three one two three two and also five day one which is another indicator under FAO custodianship so this could be let's say could be considered to improve let's say already existing data collection system so normally yesterday they mentioned also the 50 by 20 30 initiative let's say that this is a sort of continuation and expansion of the agri-survey project the 50 by 20 30 initiative let's say that upscale the efforts of of the agri-survey by bringing together two two different projects the agri-survey project and the lsms isa initiative of the war bag where lsms isa stands for a living standard measurement survey integrated the survey on agriculture and the objective of this initiative is to have by 20 30 uh the one of the like the two integrated survey implemented in 50 in 50 countries that are not currently like collecting all the agricultural information that are needed for policy making so this could be um let's say an initiative that that could be considered and if you are let's say interested to get in touch with the the team that is that is working on this initiative you can express let's say your request of technical assistance to to the FAO so before entering into the section on data disaggregation I would like to see if there are any questions on on what I've discussed so far and also if my colleagues want to to add something on this first part of the presentation there is one hand raised Mr. Campbell Islam if you allow him yeah and someone wants to ask a question yeah Mr. Campbell go ahead and yeah you have to unmute Mr. Campbell yeah I raised the hand for panelist permission not for a question I had seen that before yeah okay so it seems there is no no burning question anyway we'll have time as soon as the end you can continue either okay so um let's say I wanted to take the occasion of this this training to to discuss another topic that is very relevant in the context of the SDG indicators which is the lens aggregation so with the adoption of the 2030 agenda for sustainable development members state states have let's say embraced the principle of leaving no one behind so like let's say leave all the the the spirit of of society and in order to to reach everyone and like to make monitor progress for everyone more disaggregated data than currently available are needed so let's say that national aggregates are not enough but SDG indicators should be disaggregated by relevant dimension and this is let's say stated and represented by an overarching principle of data disaggregation that is at the core of the SDG monitoring framework that states that SDG indicators should be disaggregated were relevant by income sex, age, race, ethnicity, migratory status, disability and geographic location or other characteristics that are relevant for a specific indicator and why why this because the data could actually speak for those that are left behind and one thing that emerged with the 2030 monitoring agenda is that very often the big picture doesn't provide a full picture for example you could have a figure at the national level that hides disparities and inequalities at this up national level or for specific subgroup of the population so disaggregated data are needed like to understand who in a specific context vulnerable population are that could be like population in a specific area or specific ethnicity so to understand who these people are where they are and how many they are in addition with the SDG monitoring framework let's say the importance of formulating policies that are data driven took a greater significance the real problem is that very often those that are left behind so populations that are more vulnerable let's consider for example I don't know the migrants or indigenous people are often those for which we have less information so very often we have this data we have a different circle that policies are not formulated on specific target population because they are not available but also like the demand of data arise when some policy needs to be formulated so in the end there are always like the lack of data on population for which we would need to develop policies of course the more disaggregated statistics we have to produce the more are the challenges that national statistical systems have to face both in terms of cost of data collection and analysis but also let's say quality concerns confidentiality and sensitivity and the challenge here is that we don't only have to produce disaggregated data but we have to produce quality disaggregated data so there's the need to on one side develop new data collection tool if the information are not available but on the other side as we cannot let's say always start user base and always like collect additional data we have to use methods that allow to improve the use of already existing data sorry and also like when when you provide disaggregated data especially at this sub national level you have to ensure comparability of your time and also within countries so recognizing all the challenges posed by national data disaggregation to national statistical system the EIG SDG which is the interagency and expert group on SDG indicators has created a working group on data disaggregation that has the objective of strengthening the national capacities and develop many necessary statistical standards and tools to produce disaggregated data so this working group is composed by many custodian agencies including DFU and most member countries sorry so this the work of the the working group was created I think around 2016 and this led to the development of many tools the one of the most important and the simplest one is this minimum disaggregation set and the compilation of categories and dimensions of data disaggregation basically for each SDG indicators including also 231 and 232 the EIG SDG and the working group identified the minimum required dimension for disaggregation so the mandatory dimension for disaggregation and also the future dimension so those that are not mandatory meaning that they are not in the name of indicator or explicitly like mentioned by the target but they would be useful like for policy making then they could also be computed for specific indicators of course like in order to disaggregate data we also need standards and classification and also methodological tool so there were other resources that were developed for example a comprehensive summary of disaggregation standards and classification for SDG indicators a compilation of policy priority by disaggregation dimension and a completion of methods and tools for data disaggregation all these resources can be accessed by the link that is presented in this slide and that will be circulated after the presentation so okay here I just to explain what this compilation of categories and dimension is I have already said so the most important one is this so the minimum set of disaggregation that is the disaggregation dimension that are mandatory for an indicator so that sooner or later should be produced and this dimension are those that are specifically mentioned in the title of the indicator or in the target the other let's say okay other current disaggregation are those additional disaggregations inside the one that were one that are currently included in the database and then we have future additional disaggregation that are disaggregation dimension and category that are mentioned in the metadata for future disaggregation that are not mandatory beside that the DAGSDT has also created like within the working group on data disaggregation a task force on small area estimation so small area estimation techniques are methods that allow to combine multiple data sources to let's say increase the quality of disaggregated estimates for very small domains such as for I don't know I can bring the example of Italy Italy is divided in region provinces and then within each province you have municipalities so very often with the sample survey you would be able to the DAGSDT estimates at regional level if you want to go below this disaggregation you may need like model-based techniques such as model estimation techniques to produce accurate estimates also for example at municipality at municipality level and these approaches basically are based on the integrated use of multiple data sources such as for example a survey that is back up from census which has let's say a bigger coverage of the population and allows it to be producing well-accurated estimates okay so after this introduction let's see which are the disaggregation dimensions required for our indicators so 231 and 232 so the mandatory dimension are the sex of the old ingore household ads so when like when you will produce an indicator 231 and 232 you will have to compute them for the total let's say population but also disaggregated by the sex of the ad of the old let's say and this is already done like this disaggregation dimension is already included in the in the global indicator database then you have the type of enterprise if it so you the indicator should be computed by all farms but also by type of sector this dimension is not currently included then you are not requested for the moment to to compute the indicator at this disaggregation level mainly like I would say the main reason is the data vulnerability issue so the indicator for the moment should not be computed by this dimension then of course the size the size of the farm so the indicator should be computed both for small scale and known small small scale food producer and then the most difficult one in my opinion the indigenous status so for the indicator in theory should be computed by total population then indigenous and non-indigenous of course this information is often not not included in service so that this may be maybe challenging let's say to even by 232 to produce indicators disaggregated by this dimension and then future disaggregation are the age of the old thing that of course some age classes should be defined and then the geographic location so the the simplest one is let's say geographic locations are created by urban rural and if possible I would suggest and ask you to to produce the indicator also by this disaggregation dimension so where the old thing is located and this is because like we are planning to to add this dimension to the to the global SDG database but also because in the in the context of the new FAO strategy data disaggregated by urban rural location would be particularly relevant as a state accelerator so this would be another suggested dimension for for the disaggregation but by geographic location could be also at the sub-national level so I was saying by for example by region by province or municipality or depending on the let's say the administrative divisions of Bangladesh so okay this I have already said that so currently the indicator should be disaggregated by sex of the household and and the size of the enterprise I would let's say suggest what also the urban rural disaggregation among among this the most challenging one in terms of statistical methods to be to be adopted is the disaggregation of the sub-national level so like below let's say the biggest divisions and and let's see why this so both indicator 231 and 232 as we as we saw are most often based on sample survey data so which are the the main challenges so there are basically two big challenges when using sample survey data to achieve let's say granular disaggregation at sub-national level on one side the the sampling size let's say is often not large enough to guarantee the line of estimates for for the maximum of domains and on the other you may have let's say small area so small geographic domains that have no no sampling units so you may not have sampling units for all the possible disaggregation domains so these are the two let's say main challenges when when you work with survey data so to solve that not only for these two indicators but for all the indicators based on on survey data OCS so the office of the chief statistician of the field has produced guidelines on on data disaggregation for those as the indicators that are based on survey survey data so these guidelines offer both methodological but also practical tools in in the form of like software packages to produce direct and indirect disaggregated estimates of the SCG papers I will tell more about what is direct and what is indirect estimates and then the the publication also provides the tools to assess the the accuracy of of these estimates and present strategies to improve the accuracy of the estimate the guidelines can be accessed from from the link in these slides so and this application is particularly relevant because approximately the 30 percent of the the all the and the SCG indicators included in the global SDG like in the SDG monitoring framework are based on survey data and only for what concern the FEO 7 out of the 21 indicator under its sponsorship can be computed only or also with with survey data and they should address by the publication are those that I've mentioned so that when you use sampling and sample survey data there are limitations to produce reliable disaggregated estimates and the techniques that could address this limitation are often not mainstream in in national statistical offices so I try to to provide the content of the guideline with this idea so we saw which are the two main limitations the sampling size is not large enough to guarantee reliable estimates for all these disaggregation domains and you may also have not sample domains so these two issues can be addressed at two stages at the design stage by adopting sampling design that guarantee let's say a sampling size that is large enough in each let's say disaggregation domains of course the more you enlarge the sampling size the more the the cost of your survey will increase and this may become not sustainable for any disaggregation domain on the other hand you could address these these issues at the analysis stage by producing indirect estimates so using methods such as for example small area estimation method that cope with the little information available for these so-called small areas by borrowing strength from external data sources and these methods now like these are ideas that came to my mind while listening to two issues that you raised during this first day but this method could also somehow be used to not small area estimation methods but other data integration method could be used to solve the issue that that you have to like one one let's say the variables needed for the computation of indicators are collected collected through through different data sources so this is something that we also may discuss later when we see your data so let's see here I will go on briefly because this is let's say not the core of this training but so as I said there are two approaches addressing data disaggregation at the design stage and at the analysis stage so at the at the at the design stage like basically you have to design a sampling strategy that will ensure the presence of a sufficient number of sampling units in each disaggregation domain and this this number of sampling units will allow you to compute so-called direct estimates so estimates based on the sample observation and let's say this approach may be straightforward when the number of units belonging to a given subpopulation can be determined from the sampling frame in this case we only have to let's say choose the degree of oversampling to apply of course this will come with a cost it's more complex when let's say you have to produce disaggregated estimates for members of rare subpopulation like for example I know the migrants belong less this this type of population and and let's say that the membership of these rare subpopulation is not known in advance from the available sampling frame so from the sampling frame you have for example you are not able to see who is an homeless or who is a migrant in these cases there are many let's say sampling approach that could be used like for example I just mentioned some the large-scale screening disproportionate stratified sampling two-stage sampling or like by adopting multiple frames uh okay I will skip this in any case this slide list uh some of the let's say sampling approaches that could be used in a when when disaggregated estimates need to be produced uh did the guideline also provide some tools to Ada sorry yes we're we're very late you like you are over 10 minutes already so I spoke more than 45 minutes unfortunately yes okay I think it's also because we started late because I don't okay so thank you Faridun um let me see what I can skip okay so uh at the sampling design stage we saw let's say data disaggregation on the design stage for the analysis stage I was as I was saying data disaggregation can be addressed adopting indirect estimation approaches the guidelines develop some of them which may be useful I will not then enter into the detail but they may be useful to achieve data disaggregation at a very detailed sub-national level so uh let's say that OCS if you are interested ask the capacity to provide technical assistance on on these approaches not only for SDG ligand or 231 and 232 but also for example for 5 5a1 so with this I'm sorry if I exceeded the time and I can close the presentation the important is that I mean to to recap uh the the only thing that you have to to get from this presentation for now is that the SDG indicator 231 and 232 should be computed by sex of the head of the also by of course size of the farm so small-scale food producer and known small-scale food producer by uh rural uh urban location and in case you need support from the FAO to implement data disaggregation at very detailed sub-national level we could support you in the in the application of indirect estimation techniques such as money estimation techniques thank you very much thank you Aida for this very detailed presentation and the presentation has basically two parts one was dealing with the data sources for computing this and these two indicators and the second part was dealing with the data disaggregation which is a whole domain with all the challenges that Aida detailed particularly the techniques and the statistical method that can be used to overcome the challenges so maybe we take a few questions before we go to the next presentation uh if there is anybody who want to intervene and to ask questions to Aida on her presentation please raise your hand I see some hands yeah so who wants to start Mr. Alauddin may I start yes I have a few questions for Aida in her presentation he mentioned the under every survey this is a 10-year rounding survey every year collect data on production is it a final survey or individual sample every year okay my question is clear or not yes yeah so shall we take the other question or Aida you want to answer this okay second one is your your data desegregation you have mentioned one is type of enterprise according to type of enterprise but a household which has a crop production maybe there's some fisheries cultivation or maybe there's some forest in which criteria you um desegregate that household okay Naman can I take these first two questions and then if Mr. Alauddin has finished his question you can answer to him is that the the two questions are yes yes yes okay so please go ahead Aida okay so um for the first one so the agri-survey model is a model that has been designed but it's not a fixed model so every time it's implemented in a country this is tailored to the specific country and this applies also to the sampling design so the sampling design is not a strict let's say you know always the same one but it's uh it's designed with the country considering its specificities and also like taking into account what is already available in the country very often like countries already have an agricultural survey so the approach is to start from that existing agricultural survey and and adapt it and possibly improve it so there's no real at least in my opinion there's no real answer for your question because it really depends on on the country and the specific survey is it correct Naman I think uh yeah I think the agris the idea is like the the census you know you have this core module and the complementary module it's the same philosophy for the agris so you have a kind of core module in agris which is repeated every year and then from depending on the module you add a module on labor a module on this and on that so but you keep the the core module and the sampling is done at once yes yeah okay so in the agris we can also circulate that there's an agris handbook so where they present the methodology so there's a sampling design that is suggested for the agris survey but then I think every time the sampling design is tailored to the specific country and I think the philosophy is that but if you go to a specific country you need to look at the situation there yeah and for the second yeah yes for the second question on on data desegregation you you are perfectly right indeed that dimension is still not produced let's say the indicator is not desegregated by type of enterprise yet and there should be still some discussion with the focal point of this indicator within the FAO to decide how to go with respect to that I think that's a good question and it needs to be more discussed any other question is it I have a suggestion you know during the Amis project you know you have you know about the Amis project you know it was a project by a few hundred billion million decades so Bangladesh was one of the country and it was for two years in Bangladesh I was the lead national consultant but within that one what I did you know I proposed one integrated sampling frame so there will be a master sample then within that one they'll go for one area survey then within that one there another small sample size from the same frame for the you know you know yield and later on for forecasting or something like that or damage service something like that so if from the agriculture sensor that Bangladesh has reacted decently they can prepare one big you know master sample and from that one gradually they can add one more as the disease is doing so it will you know minimize the cost and you can leave the households and whatever units you are you know in future so as you have indicated there are seven indicators out of 21 that actually is so if you can plan something one integrated survey design more or less other countries may follow from their agriculture sensors so that will be that will minimize the cost actually in the long run and again you can relate to the indicators in between so you can also internally validate or find the relationship what is going on there so this is my proposal and at the headquarter level you can discuss this point agrees is there but agrees is module by module but sampling frame that that may be designed in that way but regarding the designation you rightly say that a smaller estimation is a good proposal otherwise sample size would be very bigger to you know to represent all the you know the wall level desegregation so I think smaller estimation is a one and widely and you know Bangladesh also use a smaller estimation for hives they've already made everything so they use the wall bank probably LLB method probably and also I personally know so what about two piece destroyers two piece destroyers that use the latest multi-level you know m-quantity method of a smaller estimation so if you just if we can you know increase the capacity of bbs so perhaps yeah that is fine you know smaller estimation should be actually promoted in here otherwise you know first will be very much higher okay good to know so perhaps we should have another discussion like on that and see if we can I don't know plan our training or yeah you need to organize specific training on the yes because the small area estimation techniques are straightforward but in application in concrete application is tricky and you need the specialist in sampling again using smaller estimation in agriculture is also very useful to know that there are some capacity there already like that they are using it for poverty mapping so I think we can continue this discussion yeah let us take another question because there are a little bit behind the schedule and then we move to the next presentation is there any other question if not I don't know okay okay thank you thank you thank you very much for your presentation to ija we we gather knowledge our some indicator sdc indicator 2.4.1 2.1.2 and 2.1 type a1 and last 2.3.1 2.3.2 can we use the same sample frame for indicator 2.3.1 number 2.3.2 for the survey sorry your question is if we can use the same sample is that the question yes can you use same sample sample for 2.3.1 2.2 is really 2.4.1 and 5.1 we can use the same sample so uh so you mentioned 2.3.1 2.3.2 5.1 and 2.1.2 I would say yes I would say so yes I think because the reference population is the same so 2.3.1 and 2.3.2 refers to either farms or agricultural households right then 5.1 refers to agricultural population so it's it's the same um then 2.4.1 I think also yes I think is also referred to to agricultural population I don't know if if Babi is still with us uh but yes I would say that the same sampling frame can be used yeah I think you answered the question right I think this as long as this all are related to agriculture yes I think we need to promote to integrate things you know running one survey designed for yes of the indicators it's not really recommended it is it's costly and it's not recommended and then as I said you know the unit of uh I know the data collection unit is agricultural household agricultural holdings for all of the the indicators that you mentioned therefore having one sampling approach for this thing should resolve the whole hassle recommend to do to do so okay I think if there is no more question we should move to the next presentation but I'm looking at the time Faridun uh how far are we from the break or shall we take some time out of the break so that we have at least this presentation before the break hello Faridun sorry I was muted we have 10 minutes to the break 15 minutes to the break like let's do the presentation yeah let's do this let's go for the presentation okay Mr. Muhammad Saddam Hussein Khan is uh already in the panelist so you can uh share your screen and share your presentation okay thank you very much thank you very much I'll do it accordingly thank you okay Faridun we're using two devices because the echo is coming at least mute one after the break I will share the presentation can you share your screen so can I can I start the presentation now or after the break yes now now yes yes now now now yeah so can I start my presentation now or after the break yes you start your presentation now please will reduce a little bit the time for the break yeah we can see the presentation please go ahead thanks for the opportunity good afternoon everybody this is Saddam from Bangladesh Board of Statistics welcoming you all in my presentation I am presenting on behalf of Bangladesh Board of Statistics I'm working as a deputy director in agriculture wing under the minister of planning and I have segregated my presentation in eight parts first one is introductory part and then I will explain regarding the survey and census that is usually conducted in BBS I will explain some of the activities those are related to the regular activities in agriculture wing and I will also explain the recent agricultural census 2019 and I will also explain the main objectives of agriculture census 2019 and the questionnaire which we use for collecting data on agricultural census I will show that questionnaire and I will explain something on it and then I also explain something related to the agricultural sample census which is recently conducted in 2020 then we're getting the sample frame that was used in agriculture sample census also domain sample size and especially I will go on focus I will go on focus for is this indicator in agriculture census and the relevance survey that we are proposing and then expectations from the FAA team basically Bangladesh Board of Statistics is a national statistical organization in our country and we have a statistics act which is called statistic act 2013 and according to the statistics acts the BBS is really responsible for collecting data generating data and generating the official statistics in Bangladesh and the main function of Bangladesh Board of Statistics to collect compile analyze and disseminate data for generating the official statistics which represent the social economic demographic sectoral data and the some of the surveys which BBS conduct first of all the household income and expenditure survey which is called HIS and then BBS conduct multiple indicator cluster survey which is called MICS and then monitoring the situation of vital statistics of Bangladesh which is called MSBSB and then sample census after the census then we usually do the sample census and we for the agriculture census we conducted the sample census and then national child labor survey and labor for survey climate change and disaster related survey we usually do and then we also do the literacy assessment survey health and demographic survey and national hygiene survey and also we conduct the food and security and nutritional surveillance survey and then the main three census BBS usually conduct population and housing census our upcoming populations and housing census will conduct upcoming October 2021 and agriculture census we conducted in 2018 the data collection was completed in 2019 and then we are also going to conduct in economic census in 2023 and last economic census was conducted in 2013 so these three main census BBS conduct and the activities is regularly done by the agriculture ruin in BBS all about the eight activities first of all the cluster survey is done in quarterly base and then cross estimation survey it is done in seasonal base and then crop forecast survey two times season based only for the major crops we usually do the crop forecast survey and we usually do the crop cutting it is usually when the you know the rice you know rice maize these are the crops is on the yield stimulating the yield we usually do the crop cutting survey and then we also do the enouale land use survey also the irrigation survey this is also enouale and labor waste survey this is conducted in monthly in every month we usually do the river waste survey and also the cross damage survey after the natural calamity we usually do the crop damage survey these eight regular surveys usually conduct by the agriculture ruin of BBS and then I will explain regarding the agricultural census in 2019 which we conducted then agricultural census held in 2019 it was in modular approach at the first stages we collected all the general information through a short questionnaire on crops fisheries and related to the livestock as well at the second stages in the sample census we information collected through a long questionnaire it is all about the eight modules we collected the information through the sample census I will later on show you the whole questionnaire of the sample census and then the which are the relevant objective we focused for conducting the agricultural census 2019 first of all the land utilization stands in terms of leasing cropping pattern irrigation status livestock population and relevant characteristics and area under fishing these are the main you know first objective of agriculture census and also we try to collect the data relate to the household which will use for the frame for the road for collecting data on the different surveys also we did the in relevant we also did the informations relate to the agricultural census that enable to observe the structural changes in the agriculture nowadays the agriculture activities are changes due to the you know repeat industrializations and due to the you know varieties of crops so after collecting the agricultural census data we may enable to observe the structural changes that occurred in the agriculture so these three in the main objectives of agriculture census 2019 and then the short questionnaire and when we conducted the agricultural census in 2019 we use the short questionnaire it's around 25 key indicators we collected through the question and I can we can show the question as well here see I'm just just a little bit small as I scan from my book see these are the question is all about 25 you know key indicators we collected data from the short questionnaire in most probably it is not showing on the screen oh we have it why we don't see it have you opened another link okay I'm just I'm just trying to you know show from the hyperlink you know oh that's why it is not coming probably you have to know come out first and then show it again you can show it later on after finishing your slides I think we all received the okay thank you I already I already sent to you through the email along with the presentation then short questionnaire we covered 25 key indicators from the short questionnaire and then we also conduct the sample census in 2020 it is all about you know eight modules sorry we cannot see the hyperlink that you open because when you share the screen you selected only the presentation so we we can only see the power point but if you want to share another file with us you can stop sharing your screen and sharing again selecting the page that you want to show okay thank you very much thank you very much so I will show you later on the agricultural census questionnaire and the sampling frame the sampling frame which we used in conducting the sample census our total innovation area was 163,672 the crops innovation area is 97,820 the livestock innovation area was 42,628 the fisheries innovation area was 15,534 and non-farm innovation area was 7,690 the categories are separated in majority based from the census innovation area these are the sampling frame we used in our sample census and the domain and the sample size the three main domains we focused which was crops fisheries and livestock and for these districts it was you know all about the 64 districts we you know covered three domains along with our you know total PSU is 9875 and total household was 148,000 and the seven module as I you know informed you before the seven module we used in our sample census questionnaire in 2020 the module one is related to the household members landowners and land utilization types module two household agriculture it's related to the crops information module three that was household based fisheries information module four household livestock the cattle information module five household poultry information modules is engagement power of household in agriculture sector and finally the module seven was the food security related experiences machinery we included there and collected that accordingly and it comes through the agriculture census linked to the ISRI indicator the five major indicator we focused on our sample census it is indicator 1.4.2 it is indicated 2.3.1 it is indicated 2.3.2 it is indicator 5.8.1 and it is indicator 2.1.2 this five indicator five is this indicator focused on our agriculture census and the proposed project that related for achieving those ISRI indicators and first of all the food security statistics survey on ISRI indicator 2.3.1 and 2.3.2 we are proposing to conduct related survey for getting information through the ISRI indicator 2.3.1 and 2.3.2 livestock and the fishery survey and preparation of food balance sheet and automation of data collection compilation and the reporting as well and then data sources which is indicated 2.3.1 and 2.3.2 as the main data sources our agriculture sample census where we related for the operating land which covers the temporary cross area permanent cross land area and current fellow land area available in our sample census and the cost of production that cover on cross fisheries and livestock but the forest data was not available on our sample census the labor hours and day related data are included in the sample census machinery and our expectation from the FAO as we are you know struggling for generating this indicator so for conducting survey that relate to this indicator we are planning to conduct the survey which relate on our proposed project so but we need the assistance and the relevant support from FAO team and this is our you know main expectations from you and your team as well we need the technical assistance as well as the further support for conducting 2.3.1 and 2.3.2 indicator related survey on the SDZ this is all about our little presentation and thank you very much and now I will show you the questionnaire yes hold on stop sharing this question was used for collecting data on agriculture census 2019 the short questionnaire it is all about 25 key indicators we collected from this questionnaire this question was used in 2019 yes you are in what format you can increase the yeah yeah we can now we can read it okay see this question was used in for conducting the agriculture census 2019 in Bangladesh see the first of all the name of the head of the household name of the father and mother occupational code and then the household is like the fisheries household the mobile number this innovation area you know the member of the household like the male female the land of household oh see these are the land under the temporary crops and then this 15 land under the permanent crops you know see do you have livestock in the score module yeah yeah in the downs I'm just showing you see the question to help the number of 40 okay number 13 the number of livestock it was given on the question as well in the production do you have production as well production is in the complementary modules I think yeah the sample modules yeah so we collected data figuring the productions from our sample census I will show you see this is the number 16 the ownership of the agriculture machinery this is number 16 like the some of the agriculture machineries are used by the households so we collected data regarding the ownership of those machineries like the tractor or tiller these are the machineries are used in our country okay that's good okay now I if you allow that I can show you the sample-sensor questionnaire okay is it okay that's this also same these two questionnaire to you know to the team so that they can when they have time they can go through you know with father you know so this is the sample census questionnaire first this part is related to the innovation area details area division restrict the administrative part of our country like small part like the union word moja the village's innovation area number the RNMO it reflects the rural municipality and the other urban area the PSU number PSU category we segregated the PSU number based on the categories like the crops fishery and the livestock as well and then household number as per the main census then household number as for the sample census you know for targeting the house we gave the two numbers household number as per the main census there was number in the main census the hot on that household and we try to keep it the same matching and giving the keeping these two numbers in the sample census but household number as per the sample census. We put two numbers in the sample census questionnaire. And then it comes to the household information, name of the household head, name of the mother, father, name of the respondent, the mobile number, NID. This is called like the NID, national identification number of the respondent relationship with the head of the household, then module one. The module one, it was all 2007 model, as I explained before, the module one, the amount ownership, use and type of land, all members of the family to 2.5. This is number one. This is 1.1, 1.1, name of the family members and type of land ownership. And it's related to the 1.4.2 and 5.8.1. There's these indicators. This is the, you know, question related to the relationship of the head of the household. So the production is module two. Yeah, this is module two, information on the agriculture grain of household, 2.1, number and ownership of the vehicle used for agriculture purpose. Okay, so I'm going to the 2.5, which is relevant to us now, 2.5, the permanent and the temporally crop calculations, 2.5.1, land of the temporally crops in the last one year and cost quantity and the value of production. This is related on today's discussion, the temporally name of the crops, you know, the different crops and the amount of the land. And this is related to the acre and the decimal, the measurement of the land, the cost, land preparation is totally related to the production cost. If you consider the production cost, so it is relevant to the production cost, the irrigation, fertilizer, wadding, hormone and other vitamins, pesticides, cutting, lifting and transportation. This is related to the production cost, including the labor as well. Okay, do you have the same for livestock? Yes. And then the total is spent labor hours. So, let's see over here the total productions amount kg of the number and the price in taka and total by productions value. Okay, you have labor hours. Okay. Yes, yes, yes, labor hours for considering the day so we can convert the labor day from the labor hours as well. And also the by-product, we can calculate the bulk product, relevant information we collected from the sample census. Then this is all about the temporarily crops. This is all about the crops. This is all about the crops. This is all about the crops you can see here. And this is the visitable type, temporarily crops, temporarily crops, temporarily crops. This is oil type, food type. I think this is a very rich. This is related to the fiber, sugar and the energy type crops, like jute, sugar cane, tobacco. And then temporary crops like flour type, the different flowers. So the same is available for livestock, fishery and forestry? Yes, I can show you. This is the land permanent crops for the permanent crops. We, it's a 2.5.2 the land of permanent crops in the last one year, and relevant cost quantity and the value of productions. So see, this is the permanent crops, like the mango, berry, jackfruit, leaches, the guava, coconut, juby, and hot palm, olive, dates. These are the permanent crops, plum tree, custard, seed. This is the kind of energy type, the betel leaf, betel nut, tea, wood and the forest type, subamble, libeck, ventri, mahogany, teak. This is 2.6 sale and the extraction of the agriculture resources. The module 3 cultivated the fisheries information in household. This is module 3. This is related to the fisheries information from that household. Is that aquaculture or is it the fishery, inland fishery? This is inland fisheries, land fisheries, land fishery, the land use and quantity under the fish farming. And 3.2 is the fish farming, it's the aquaculture base, land use and the quantity under the fish farming. Okay, where is the fish farming? The cultivation in the pond and creek and diggy, land use and the quantity under the fish farming, 3.2, cultivation in pond, 3.3.1, collecting renal pona from open water source and heating and selling, 3.3.2. Fisheries cultivation and collecting, 3.4.1 fisheries cultivation under the operated land. This is 3.4.1, name of the fishes and then quantity price, last one year production, sold, which is for use for family purpose, that is heating, profit and loss as well. 3.4.1 fisheries cultivation under operated land, fisheries cultivation under operated land, 3.4.1. 3.4.2 has any member of your household called fish in open water in one year. This is related to the open air fishing. And then renal pona, this is kind of aquaculture cultivation, the renal. And 3.7 fish food related expenses. And 3.8 the fish care related expenses, module four, the cattle related. Do you have any cattle and poultry or household run cattle? And any poultry farm? This is the first question, this is 4.1 in this module, then 4.2 use of land for cattle. And then cattle of household related 4.3, kind of cows, you know, different type of cows, this is native cows. Yeah, PNCore ships, all those. And then kind of goats. And then in the right side, you can see the ships, kind of ships, you know, see, kind of goats, kind of ships. This is pig related, you know, pig related. And 4.4 purchase of sale of cattle, birth date number and the prices. And the different cattails we collected the formation from the different cattails. And this is all about the purchase and sale of cattle, and their status related to the birth and date as well, and their price. And 4.4 information, daily productions in the household, kind of cattle, the different cattle, and the number of current milking cattle daily produce milk and liter, and every liter ever is milk price from the different cattails. And 4.6 cattle feed, health and other costs. This is for related cost. And 4.6.2 is the health related cost. And 4.7 price of last year's by products. Cow, buffalo, goat, sheep, and other domestic animal. And module five, poultry of household. Are there any poultry, quail, python, turkey, etc. are available in that household. And then 5.2 years of land under poultry farming, poultry number, poultry price, and poultry related information, native cock, chicken, balanced stock, layer, boiler, chonally, farming, turkey, dog, quail, this is a different type of poultry. We usually, you know, nourishment available in our country. 5.4 poultry, egg production related status, the native chicken, the dark layer, these are the different, you know, varieties of poultry. So the egg production from those poultry, we collected the data from those. And then 5.5 poultry feed related cost for, you know, for the poultry farming, the relevant cost of food for feeding those poultry. We collected data for relevant cost on feeding and also the relevant cost of medicine and other shade related infrastructural cost over there. We collected data based on those shades and the medicine relevant cost. And then 5.7 price on last year dogs and chicken by product. You know, this is the cost for the last year dog and the chicken by product. This is like the domestic cock and others, dog and others. Module six, manpower, module six for manpower generally engaged in agriculture work in household. Manpower usually engaged in agriculture work, population type. This is the year segment, five to nine years, 10 to 14 years, 15 to 17 years, 18 to 24 years, 25 years and above. The manpower, those were engaged in agriculture work, male, female, we segregated the data, male and female also. And then some of those are involved in cropping, those are fisheries, those are involved in livestock and some of them, those are engaged in both and two or more. So we also collected those data. At 6.2, the number of laborers in agriculture work in the household is permanent temporary total. And then module seven, household food security. This is related to the SDG 2.1.2. We'll now ask some questions about your food taking. We will answer listening. The question 7.1, see, this is around seven, eight question all total, 7.1 to 7.8. It's related to the household food security. And it is focusing on the SDG 2.1.2. See these questions. Okay. It's a very detailed questionnaire. So this is all about our sample census portionary. Thank you very much. Thanks for the patient hearing. And any question you are most welcome. So our team who is over here, we'll try to supplement your portion if you have. Thank you very much. I think Mr. Sadam for this very interesting and detailed presentation of the surveys conducted by BBA, focusing on the the census, which is really implemented, I can see using the FAO recommendations. You are really going by the book and applying all the recommendations of FAO in this census. So I think we are a little bit behind the time, but maybe we'll take some time from the break. I think this was a very interesting presentation we needed to go into questions and see. Faridou, where are we with the time? So we can go to break for 10 minutes. I don't know if it's possible. And we can come back at 4.05 or we can have 15 minutes and we can come back at 4.10. All right. So we can have some questions now and go to the break after 10 minutes or not. We have one question, Mr. Taitul Islam, if you have a question. I have already got my answer. I have already got my answer on this question. The questioner is very nice. I think all the data come from we can calculate it at 2.3.1 and 2.3.2 from this question is the questioner. It is all the data I think all the data is available from this question. I have got this. Just a few minutes. Some little question to Mr. Sattam. What are the use of the residual products? Five products of the, is this question incorporate those by-products of the agriculture produces as Mr. Taitul was asking the same question. It is there. In that case, it is complete on if the definition permits. We have to check the definition for proven treatment one and proven treatment. If a farmer produce rice, then he can also produce the straw. The straw. Can we get this by-product from this rice like that other crops? Is it okay or available? Yes, it is okay. If by-products comes from the farmers, if they produce the rice, the by-product will come automatically. We also collected that relate to the by-product cost as well. This data we will get from the sample sensor. I think I will be able to highlight some problem. During the revenue calculation, crop used for rain or something like that, there were some information during calculation of the revenue. So that part may be missing from this one. It is very straightforward about the costs. But there may be some share cropping, there may be some exchange of crops. Then I don't know. I'd better check this question. Yes, perhaps it would be very useful if the questioner can be shared with us. So we have a look after the training. Saddam, I will request you to share this questionnaire with the team. Okay, yesterday we sent the presentation along with the two questioners also sent. Both of them? He shared the questionnaire with all. So any of the questions before Yakov? Yes. I just a comment. I think the questionnaire is really very, very exhaustive to make in order to calculate. I mean, I have seen different countries' questionnaire and these details of variables maybe we found in a very few countries. Therefore, in my opinion, there are some issues that have to be addressed. For example, my quick thing is like labor on livestock thing is not there. I can't see meat production. I can't see that type of things. And I mean, there are few things we need to fix within the existing questionnaire so that we will be able to compute the indicator. Otherwise, I think the main important things are really exhaustively there in this questionnaire, the way I saw it. The only thing is to sit down, look at the details of the questionnaire and organize. Now the critical problem is how can we really organize our data to have that format that we showed you in the first day and the first and second day. That's the critical thing and which is doable. To be honest, this is really doable. But the details are there except few things. And as we have been saying, we don't want to recommend to launch a new survey, a new thing, a new approach. Because the existing system with a minor fixing, we should be able to produce this indicator as a way I see the questionnaire. Okay, yes. Any other question? I have also a few, but I wait for any other question. If not, first question, I understand that the core module which was conducted in 2019 has been done by complete enumeration, right? Yes. Okay, so that's exactly what is recommended. So and then you took a sample and conducted the seven modules on that sample in the 2020. Yes. Okay, so that's number one. Now I have seen also that by doing so, you took the pre-question to put the ID of the holding in the 2019 into the sample. So we can link the two information from the core and from the sample, which is very good also because now we can have the inventory from the complete enumeration and then we can link that to the sample because the ID reflects both. Is that correct? Yes. Yeah. Okay, I think this is certainly comprehensive. The limitation is that this kind of census is conducted every maybe every 10 years. So we may be able to compute all the indicators for one year, but the issue is how we can move now if you want to do the monitoring frequently. How frequently will be conducted this kind of investigation? If it is taking too much time, we should start looking at also additional sources of information to compute the indicators between the two censuses. Otherwise, for some kind of baseline computation, I think this is completely almost exhaustive and we have almost all the variables available. Now, maybe when we see the second presentation, we'll talk also about the household income expenditure survey. We can see if in between two censuses, we can find a survey or modify the survey in order to be able to compute the indicators between the two censuses. And my last comment would be that exactly as Jacob was saying, I think after this training, I see immediately one immediate action could be to work together between your team and the team in FEO and start computing this indicator out of the information from the sample census and the core census. So that would be one recommendation that I have if you agree with that. Yes, we are totally agree on it and very well said Nehman. So really appreciate and we'll work together. Okay. Thank you. Okay, so any other comment intervention? We need to go into the detail of this question as you know this was very good presentation that we need to look at the question as in details. And as I said, after this training, we can keep in contact and start looking at the data as Jacob was saying, how this data could be organized and put in a format in a way that you can compute the indicators easily. Okay. If there is no additional question or comment, let's take a break. Faridun? Nehman, I think you have an announcement, yes, about the picture. Yes, exactly. I was going to do that. So we'll make a break now. It was requested yesterday that at the end of this training, we have a group photo. So we've been struggling to find a technical solution for that. And maybe Faridun, you can quickly explain. At the end of the meeting at 17 hours in Bangladesh, we will take five minutes in order to do what Faridun is going to explain now to take the group photo. Faridun, please. Thank you, Nehman. I just wanted to inform that I am about to send you another link, but not for the webinar, but for the meeting. And I'm back. Okay. Are we all back so that we can start the last presentation? Shalya, if you are ready, please start your presentation. Ms. Kattu. Shalya, are you there? Shalya, you may start your presentation now. Thank you for giving me the opportunity. This is the four points slide. Use this slide show, Shalya. Oh, yes. Can you compare me? Yes. Okay. Welcome to the presentation on Data Gap for Computing and Recording on L152.3.1 and 2.3.2. I am Shalya, working in agriculture in Bangladesh. Now, our topic is related to SDG goal two, where we can find the goal is N hunger for security and improved nutrition and promote sustainable agriculture. Under goal two, there is a target 2.3 by 2030 double agriculture productivity and income of small scale food producers, in particular women, indigenous people, family farmers, mass produce, teachers, including food security and equal access to land, other productive resources and enforced knowledge, financial services, markets and opportunities for daily education and non-farm employment. Under target 2.3, there are two indicators, indicators 2.3.1, that is volume of production by labor units by process of farming, hospital, forestry, and the side side. And the indicator 2.3.2, average income of small scale food producers, by sex and indigenous teachers. For computing, 2.3.1 means volume of agricultural production, that is from crops, livestock, fisheries, and forestry, and funnels themselves right, number of labor days, and number of small scale food producers. We see that there are three criteria for computing small scale food producers. Number one criteria is compute the operation length by coming the size, that is in sectors of land consummated with permanent and temporary crops and settlements. And then 15 will be included and then the house will be excluded. And the second option is hard size, that is converting the number of livestock in stock for the basis in tropical livestock image. And some of these intermediate variables will be used by species to determine the total number of species. And the third option is from the total revenue. The total revenue from all agricultural activities given to a food producer, that is from livestock, fisheries, and forestry. If some of these components are not available, it should be flagged in the indicator metadata or in the methodological notes. And then some of these revenues, and after self listing the revenue, we should make the threshold, that is the forest storage, holding, threshold identification, small holders, which is set to the... Are you moving your slide? No, 2.2, so I am now in 2.3.1. I am talking about the three types of threshold for calculating small scale farmers. So from intersection of three thresholds, we should take only the joint list that is defined as three criteria. After that, we should take the volume of production and decided by the number of small scale farmers in the formula. Labor productivity is the agricultural output divided by labor input, that is total revenue divided by labor day work. That's a gap for 2.3.1, which is the number of labor days that is absent in the household income and the expenditure survey. But it will be available in agricultural sample centres 2020. And another one is, forestry production from household is not available in agricultural sample centres 2020, but it is also available in household income and expenditure survey. So now I am going to 2.3.2. For computing 2.3.2, we need data on volume of agricultural production, that is crop, livestock, fishery, and forestry. Then we need constant sale price, revenue, cost of production, and income. And finally the number of small scale food producers. For computing 2.3.2, we need data on cost of production from all the units and all the items at the time of production data. But you see we have cost of production service in several times for several crops, not at the time. And these cost of production data we will get from the agricultural sample centres 2020. So we think agricultural sample centres 2020 will meet our data requirements. And another one is updated CLU conversion table is needed. And updated TPP is needed. And the pastoralist information is also absent in our data. And agriculture, possible data sources, possible data sources maybe agriculture sample centres 2020 and the upcoming eight years that is 2022. And we hope all that are required to come from a single survey. And the last one is conducting a company in the agricultural service is for people. In current situation, currently we don't have baseline information of a 2.3.1 and 2.3.2. Also we don't have a standalone service for 2.3.1 and 2.3.2. We know that AFL has estimated this LPG 2.3.1 and 2.3.2 using the household income and expenditure service of 2010. And the thresholds are 1 hectare for land size, 1.6 CLU for livestock and $2.3 to TPP dollar for revenue. And our parent attempt is using household income and expenditure service. We can try to compute this indicator as a proxy indicator. These are the household income expenditure survey module in 2016. That is, we have seen that there are information about operating land. We have crop production. We have livestock and poultry production. And fish farming, fish capture also farm forestry. And there are some information of cost of production also. Maybe it is not complete, but the problem is the labor days are captured in this data set. So this may not be the national estimate, but we can use this as a proxy for comparing the indicators. So I have a very brief presentation about the data gap. If you have any comments or suggestions or questions, please continue. Okay. Thank you very much, Ms. Katoom, for this very clear presentation. And now we have exactly, so I propose we combine the discussion on two things. First of all, the comments on the presentation. But since this is the last presentation, if we can discuss a little bit what could be really the next steps. And after this meeting, what could we do together in order to start computing the indicators and fill the gaps with the assistance of FAO collaboration between FAO and the BBS in order to move one step ahead. So first, we can start looking at if there is any comment or clarification to ask to Ms. Katoom on the presentation. Please, the floor to the participants. In the meantime, I have a clarification. If I understood very well, this household income expenditure survey is conducted every five years, more or less. Yes. Because I have questions. I have questions. In the slide, here is, I'll say it using the HESA 2010 data. They have calculated SDZ 2.3.1 and 2.3.2, how it is possible. There is no information available about production cost of agriculture product and number of days work against the product. This data is absent in the HESA 30 days. How it is possible to calculate this, I cannot understand. Yeah. So the question is how FAO was able to derive 2.3.1 and 2.3.2 for 2010 missing data. So basically, it means that there was someone from BBS involved in that calculation? No. I know it from presentation of FAO. There is a resource paper on this topic that FAO has calculated this as SDZ 2.3.1 and 2.3.2 from 2010 HES as top data. So I know it from, but I don't know if there is anyone involved in this process from BBS or not. Aida, do you know anything about that? Is there some metadata explaining how this was done by FAO? Hello. I don't know if Aida is around. Anyway, we can investigate that because maybe we cannot reply immediately. To me, this work has to be done mostly between Naman. Yes? I think that information is not about 2.3.1 and 2.3.2. It is on a small scale, identification of small scale farmers. Just identifying only the small scale farmers, not computing the indicators. It was part of your representation, if you remember. Can I share my screen? I think in my presentation, yeah. Okay, good. Can you stop sharing your screen? It is stopped. So I think it is, can you see my screen? Yes. Yeah, it's clear. Yeah, I think it is sure. It is referring to this slide, Naman. You see Bangladesh? Yeah, but this is the income, you know? Yeah, it's not very... It's 2.3.2. Yeah, averaging comes on a small scale for this as PPP per year. It's 2.3.2. Anyway, I see Arbab also, maybe Arbab knows something about that. Yeah, I can mention the paper that is FAO Statistics, Working Papers, she used 18 for 18, Macrology for Computing and Monitoring the Sustainable Development Goal Indication. I see. Yeah, yeah. So this is the reference that we share. If you look at the agenda, it is the same. This is the information. Can you see? Yeah, yeah. No, no, we see. Because this is a document which is being revised now, but I don't know. Anyone from FAO want to clarify a little bit on this? No, I have some problems with her microphone and she wrote in the chat. It was computed by Rulis only for 2.3.2 and 2.3.1 without involvement of PBS. Yeah, yeah. Okay. Anyway, what we can say at this stage, we could try to understand... Probably the meeting that they somehow managed from other sources and they incorporated in the calculation. Maybe, maybe the labor days they have taken. Yeah, yeah. Maybe from the labor surveys some ever is used. So in that way they calculated this. Maybe they can use other survey data. Not only, they can not only use his data, only his data, using only his data, they cannot estimate the SBJ indicator 2.3.1 and 2. Only the... Yeah, yeah. I agree, Mr. Tahid. Probably, you know, the missing data was incorporated from other sources. Other sources, yeah. From labor support. Again, you know, we are all statisticians. This statistical work, you will need at some point anyway to fill in some gaps. You need to see what is the impact of that. Because it is very rare in a country to have exactly all the information needed. So you need to use some kind of a common sense and using statistical techniques sometimes to try to get as close as possible to what you want. We make all the effort to collect exactly the data which is needed. But you know, it's so complex and you may end up having a very, very complicated survey. So you need sometimes... The simple example is the price. As Ida was mentioning, we talk about constant price. But usually you have... You miss this data. So you need to take the price that you find and try to work on this price to get as close as possible to the constant price. So the rule is a database in the send which is available in F.A.O. And they take information from LSMS data, from over data, and then try to work out something and probably that's what they did for Bangladesh. Now my proposal would be from now on. We know that some data is available in Bangladesh to work together. So this work should be done between F.A.O. and the country working together. So that is very, very clear in the metadata section. We explain exactly what is the methodology, what are the missing data, and what is... What approximation has been done in order to come up with the numbers? Okay, so that's for what I just wanted to say for the moment. Any other comment on the presentation? Naman, maybe just a quick question. Maybe this is referring to Saddam's presentation, earlier presentation. We saw all the agricultural sensors, the core modules as well as the sample thing. And he also mentioned about the annual agricultural survey type of thing, production, crop cutting, whatever. So if we hear a little bit more on that survey, because that is annual thing, is there anything that we can really incorporate in that annual survey to get some of the variables? So if you can hear a little bit more on that annual regular survey on agriculture, it might really fill the gaps, because there is the household income expenditure is every five years. The census is every 10 years. There is one survey, annual agricultural survey on crop, which is even crop cutting for major crops. So if you can, I mean, if you can see a little bit, if there is any linkage that can be made between these things. I think this question is for all participants, particularly to Saddam and Mrs. Khatum, but also anybody from DBS. We have seen that the ag census, more or less, is about every 10 years. The income expenditure survey is about every five years. So already that's better than this. Anyway, the frequency is better than the census frequency. But the census is more complete and more exhaustive. Now we have seen also from the first presentation that there is annual, the regular surveys by the agriculture wing, which include crop cutting, crop estimation survey. And we understand that the crop cutting is done on the major crops. And the farmer estimation is used for the minor crops. So is there any scope here so that in between the census and the household income expenditure survey, some variables could be included in that annual survey to get closer to the need, that requirement for the 2-3-1 and 2-3-2. Allotment may be also this question. Thank you, Mr. Noman. I think it is difficult to include because our regular survey, which for estimating the production of different crops, major crops and minor crops, these are only for we interviewed the farmers only for area, cultivated area for that crop and the production of that crop. Okay. And the average yield rate for just like simple question from this regular seasonal crops and estimation and this survey conducted after harvesting each crop. So it is a simple form and a simple questionnaire. So it is difficult to include the question related to 2.3.1 for every crop and every survey. Would it be possible at least to expand a little bit for them to ask the farmer how many cattle and the livestock part, the inventory and the products or that would be overburdening the survey? Actually, we collect livestock and faulty data from the secondary sources. It is from the Department of Livestock. We collected data from that from that offices, but we did not go to the, in regular, we did not go to the household for livestock or faulty or fisheries. Okay. Now, I was thinking in the future, if the BBS can work with FAO to see if what could be the implications of adding a few more questions? What would be the logistical cost and the training cost, etc. And is this something feasible or not? Is that, can we have a discussion on this possibility or not? Yeah, we can discuss. Actually, you know, we have a series of training from at 27 or 28 Zoom with your FAO team, first 2.1.1 and 2.1.2, which is conducted by Mr. Abdur Sattar Mondal and his team. Second was this 2.4.1 and Mr. Arbab Khan is maybe here. And third one is with Indicator 5A1, lead by Margarita Guerrero, and last this one, lead by you. And during talking with 2.4.1, Indicator Mr. Arbab Khan and his team proposed us to seek a few assistance to conduct a standalone survey for 2.1. Then we are planning to write a letter to your director as to sixth division and also talked with this Margarita Guerrero and she also advised us to write and we had, we had a plan to integrate a survey on 2.4.1 and 5.3.1. Maybe in the year of 2022 or first of 2023, maybe we can go for a survey. If your team or you suggest for 2.3.1 and 2.1, you have already seen our central census questionnaire. We have collected data in 2020. It is last of November and December. So we are in the process of data entry because it is a manual entry, long used question here. And due to pandemic COVID-19, we are sometime we have lost due to lockdown of our country, we have lost some time and we are now according to our work plan, we are some behind our work plan. But what you desire to conduct 2.3.1 and 3.2 mass survey. If you suggest to make a standalone survey next, if you assist, you can do. Or otherwise, if it is possible to include in 2.4.1, maybe in 2.4.1, there are 11 module is there. I don't know Mr. Rob Khan will suggest or not. Can we include the questions in module 1 and 2 in 2.4.1. So whatever you suggest and if you assist us and guide us, we can conduct an integrated survey on this four indicator for a standalone survey for 2.3.1 and 2.3.1 and 2.4.1 and 5.1. If I may. Can you hear me properly, Naman? I just wanted to confirm. Okay. So thank you, Mr. Alauddin for the update that you gave us on the previous trainings on the STG indicators that were given to BBS by FAO team. Now, Naman, just to give you some perspective, 2.4.1 is an indicator that cut across the three dimensions, right? Economic, social and environmental. And hence, as per our discussions with BBS, I mean, it was in fact, there was this mutual consensus amongst us that us integrating the questions from the survey module of STG 2.4.1 in the current agricultural survey will make it too lengthy for it to be administered, right? And hence, you know, lengthy survey means more cost and respondent fatigue, etc. So keeping in view all these considerations, you know, it was, in fact, suggested by BBS as to whether we can have a standalone survey for 2.4.1. Now, in the 2.4.1, we have. Excuse me, when you say standalone, it means that you conduct it once or you keep conducting it every year. This is not going to be conducted every year because the periodicity of 2.4.1 is three years. Okay, so once every three or every four years. Yes, that would be the plan. So when we say a standalone survey for STG 2.4.1, the reporting cycle for 2.4.1 or the periodicity of the indicator is every three years. So we then expect this survey to be repeated or to be administered every three years. Now, Mr. Aloudin mentioned as to whether we can use that survey for 2.3.1, 2.3.2 and 5.1. And I would say yes, because some of the information that is required for STG 2.4.1, especially the sub-indicator on land productivity, we are collecting information on the output value or value of production of the agriculture holding. And hence, this is information we very much needed for 2.3.1 and 2.3.2 as well as for the numerator of the indicator is concerned. Now, there needs to be a few tweaks and adjustments because the scope of 2.4.1 is only confined to crops and livestock. And for 2.3.1, 2.3.2, I mean, we are talking broadly about agriculture and hence we include fisheries and forestry as well. So from this perspective, there are certain things within that survey that needs to be modified for it to be able to collect information for both these indicators. Now, just for the sake of Mr. Aloudin, 5.1.1 is also one of the sub-indicator within social dimension of 2.4.1. And hence, with a minor tweak and adjustments, we can collect information on 5.1.1 as well. So with this standalone survey, if BBS agrees and then they decide to have it on regular basis, let's say for example every three years, then using the survey, we can collect information on four farm survey-based SDG indicators, 2.3.1, 2.3.2, 2.4.1, and 5.1.1. So if this is the idea then, of course, I mean, I would suggest that while designing this standalone survey, then we should perhaps have a joint discussion as to how to build in the requirement of 2.3.1, 2.3.2, and 5.1. Okay, thank you, Arbab. At this point, I don't know, but I think FAO should do some kind of homework because if all of us go to the country and ask to conduct a survey for a specific indicator, I think that will not fly. This will not be practical. Probably there is a need within FAO to go through all these indicators. I was thinking that the agrees and the 50 by 30 were doing that because they were focusing on the indicators to see if, to facilitate the work in order to, if there is a need to have a survey, to make it as much as possible SDG-friendly, covering as much SDGs as possible. And models could be the IGRIS or LSMS ISA, but I think there is some discussion within FAO in my view to try to facilitate the work at country level. But that's just my, as an external consultant, that's my view. And, but I see, Amirul, you have something to also. Yeah, I just want to add with what Arbab just said, because Arbab rightfully said that whatever the surveys that are conducted by PBS, the module is already big. So if we try to add this one, we have two problems. One is something design problem, another one is load, extra load to the interview at the same time interviewer. So one standalone survey incorporating these four indicators, if we can plan, that can relate to my earlier suggestion that one big sampling frame for 2.3.1. And then from that one, we can just select another sample that covers the restricted size of the, the restricted definition of the 2.4.1 and 5.8.1. In the process, we'll just adjust the sampling weights. And that will, I think, you know, that one standalone survey, initially incorporating the coverage of forestry feature is something like that. And then when you can, at least for 2.3.1, 2.3.2, and when you go for 2.4.1, from that survey, we'll all reduce the sample size, but at the same time, they adjust the weight to represent the national level. So by this one, one survey, standard survey should serve the purpose. But now, as you said, rightfully, a few eight-quarter can, you know, the unit lead may discuss among themselves how better it can be treated. It will reduce the burden, the resource burden for every country, if you can do so. Thank you very much. Okay, yeah, the mass sampling frame is one up. But Yakov, you want to say something? Just a minor part. I think this is, you know, we need to discuss this really further, you know, the FAO and the countries. I think in general, the FAO standpoint is not to bring something new, to build on the existing system. That's what we are really promoting as much as possible, instead of like, you know, recommending with this thing, like standalone surveys for, you know, two, three indicators together, you know, all those things are really, really good in theory, even for each indicator, that would have been very ideal to be honest. But we need to make sure that, you know, especially the resource, where are we getting these resources to do those type of things? Okay, and if it is like, you know, if you can have a special project for each of these things with a resource that FAO is providing to the countries, I can understand that can be easily implemented. But, you know, in most of the cases, these type of things, I mean, I mean, in order to make it sustainable, continuous, you know, countries have to own it. Therefore, I think in as much as possible, we need to rely on the, how to really improve the existing system in the country, so that we'll be able to produce the required indicators. That is why we are really starting to understand what I know this discussion this morning really helps us to understand, you know, what's coming up from a household consumption survey, what's coming up from the cultural census, and, you know, we are investigating if there is something we can really, you know, link with the annual agricultural service so that, you know, within the existing system to produce some something which are really important to address these data needs. I mean, that is, you know, we need, I don't think we need to finish this discussion by today, but, you know, this is the type of things we need to brainstorm. Yeah, okay. That's good. Are you still metered or are you, now you can? Hello, can you hear me? Yes, now, yes. Ah, finally, okay. So, no, I fully agree with Jacob's comments. I think this this need further discussion and possible we should try to use what is what is already there instead of designing new data, ad hoc data collection system. So, perhaps we really need to coordinate also with Margarita, with Arbab to see how we can join efforts. Yeah, in a way that is sustainable for the country, because, yeah, we cannot design a survey only to compute some indicators. So, yeah. Okay, so I think concretely, first of all, there are very positive things in Bangladesh. One, we have this census of agriculture. Now, I understand that the processing is not complete, but I hope in coming months, the 2020 sampling part of it will be processed and the data will be available. I understand also that the for the core census, at least the data should be available from the 2019. And this is a very rich source of data, probably at least to get some kind of baseline. I think this information would be in my view, very useful in starting the process. So, that could be started immediately because the core data is available and the team is working on the sample data and the FAO team can liaise with Bangladesh to start working on that. We also have the household income expenditure survey. Even if it's not annual, but at least the frequency is every five years and it's also somehow good. And this can be another complementary source of data that can be used to start compiling the indicators. Now, for the future now, I think there should be, in my view, discussion within the FAO and the countries, how we really want to assist the countries in putting in place, as I think Jacob or Ida was saying, some sustainable way of getting the data needed for the indicators. So, to see what are the different approaches, Amirul was talking about the mass sampling frame approach or the agrees with the modular approach, etc. But first of all, to see, to put all the indicators on the table and see what are the requirements and what are the indicators that can be joined in forces in a way in order to see existing surveys, how can we add some variables in order to cover as much as possible those indicators. But to me, an action plan could be immediately to start working with Bangladesh on the available data to go through the process of computing 231 and 232, particularly using the agriculture sciences data and the household income expenditure survey. And then in the future to maintain the dialogue between FAO and Bangladesh, because there are many, many indicators coming, we all come and do training and we want the data for our indicators. So, how can we make the work easier for BBS? So, there should be some discussion within FAO and discussion between FAO and Bangladesh to come up with a plan which is sustainable to fill the gap in terms of data. Is that something which makes sense? Yeah, I think at this point, this is a good proposition because after this session, our communication will be still alive back and forth from the headquarters and BBS. You can discuss it yourself. And first, you can look at the questionnaire of agricultural sensors that is shared by Mr. Sattran. From your side, just look at it if it is possible to create indicators 231, 232 for this year. And then if it is for this baseline, then future service may be planned or integrated with other regular service. That in that case, you can also suggest in which survey will be best to add this module. So, I think, side by side, BBS is writing, I think so far I understand, BBS is writing to the agrees to get support from them to report on 2.4.1 and 5.8.1. But at the same time, I think it can also be considered if 2.3 and 3.1, 3.2 can be considered. So, this is why BBS is not capable of proceeding further on this indicator due to the COVID shock and something like that. But BBS has now, as we have given the training, BBS is now well-equipped with this idea and then calculation process. So, with little help from headquarters, I think BBS will be able to do that one. Let BBS write to agrees on these two. Initially, with Arbab, we discussed it, and also Margarita suggested to incorporate these two. Let's see what response coming from agrees are also the headquarters. And by this time, you three team work together. If it is feasible to club these two indicator to that one, you can also plan that one. But without your support, not only a technical, also some research support, BBS probably not in a position to conduct immediately some of the indicators, because a lot of the proposal they have given to the government, it is waiting for their response. So, we need your support for the time being. And once it is on track, then subsequent surveys may be supported by the government. That is the best I can say for the time being. Thank you. Thank you, Dr. Amirul. I think Arbab wanted to also comment. Arbab? Yes. On this very point, you mentioned that we should have an all-encompassing 360-degree strategy as for internally FAO is concerned, while once we reach out to countries for providing them support on the SDG indicators that we are custodian of. Then from this perspective, while we were discussing 241, we discussed agri-serve program and 50x2030 initiative. Now, Naman, just remember that in both these initiatives, the needs of 231, 232, 241 and 551 are already built in. So, these are already reflected in the questionnaire of both these two initiatives. Now, Bangladesh being eligible country for the 50x2030 program, what we suggested to them was to write to the 50x2030 team, because sooner or later FAO will extend support to Bangladesh in terms of upgrading or improving their agriculture statistical systems. So, from this perspective, in the longer term, FAO has a strategy for improvement of agriculture statistical systems of developing countries, of which one is Bangladesh, an eligible country. And hence, we think of that as a long-term strategy for Bangladesh to make its reporting systems more robust as for agriculture statistics. Long time is how many years or how many, when you say long time? Exactly. So, this is a very pertinent question. And hence, while we were discussing this within the context of 241, I mentioned it to BBS that please write to the 50x2030 team immediately, so that they can reward to you and we can come back to you as FAO with a sort of timeframe as to when we can support you in terms of improving your systems. So, that's something which is not in our, as being focal point of individual STG, that is not within our mandate, but it is certainly within the mandate of the 50x2030 team and Agra Service Program. So, they can better reflect on the timeframe as well as the commitment. Yeah, but as focal point, I think we should also advise the management to have a dialogue with the country as an FAO or as an FAO statistics division or whatever, so that I think this is a very good perspective if the 50x2030 can be, if Bangladesh is interested and make a request, and this can be rolled out in Bangladesh as soon as possible. To me, that will solve a lot of problems. Instead of each one of us as focal point going and asking for a single standalone survey, this would be a more strategic approach to filling the data requirement for the computation of the indicators. So, as far as I think for our training, probably I would suggest that we make a recommendation in this sense if you all agree for FAO to consider that data requirement so that the 50x2030, of course the request has to come from BBS if the Bangladesh is interested, but to look at this more integrated approach of working with the countries. But that could be really the solution in my view to avoid this piecemeal approach of each indicator is going to the country to have a dedicated survey. That would be very costly and I'm not sure, maybe Bangladesh, but I'm not sure many countries will be able to support that kind of surveys. But anyway, so we are now really in the core of the what we do next after this meeting. So, if you think that this could be a good approach, we can stop the technical discussions here and then we try to summarize or to wrap up the meeting and after that we have the group photo before closing the meeting. But maybe we cannot exhaust all the discussion on this topic now, but let's keep because we have to enter to that request for the photo session, we have to leave this session. So, during the photo session, we can save the words if we want to before we close. Yeah. So, but before closing, I just want I was taking some notes and I will show this in the screen as possible conclusions and next steps for this. And I'm on one question. And do you suggest that you know, VBS also write 2.3.1, 2.3.2 in the request that is being sent to EGRIS along with 2.4.1 and 5.1. Should they include this indicator in their request? I don't know, Mr. Al-Abdin, how should we go about this? Excellent, Mr. Amirul. But we are seeking your suggestion. Can we write or we can only write for 2.4.1 and 5.1.1 and later on 2.3.1 and 3.2. We need your suggestion. We can include these four indicators. So, what would be your suggestion? So my suggestion would be that, you know, VBS write to FAO and, you know, copy all of us, right? The focal points and write to Christof and Flavio who are representing. I think they should write to Pietro and to Jose. Perfect. They should target the highest level for decision making. This is a strategic decision. No problem. So that's even better. So and as for EGRIS survey program on 50 by 2030 is concerned, I mean by default, all what I was trying to say earlier that in their questionnaires, in their modules, all the needs and requirements of all these four indicators have already been captured and integrated. So once FAO comments technical support under 50 by 2030 or EGRIS survey program, by default, Bangladesh will be ready to report on all these four farm survey based SDG indicators. So from this perspective, I mean even without, even if VBS doesn't mention these four SDG indicators, I mean they will get SDG indicators in general. They need support of FAO in order to, so we can take that and look at the details. But I would say that Mr. Aloudin, if VBS can certainly go through the FAO office in Bangladesh, so Mr. Amroul would be also looking at that. So that an official request is sent by Bangladesh to the management of statistics units in FAO to help Bangladesh in dealing with the SDG indicators, something like in this line. Okay. May we add to Mr. Joseph Rosario, Rosario director of statistics or Mr. Fetro Ghaneri? You can write to both of them. In my view, I don't know. Both of them. You write both of them and copy FAO Bangladesh. Yes. Yeah, FAO Bangladesh should be. We can write one of them, then I can go copy to others or individual letter to director and chief statistician. No, you can write to both of them in my view. Okay. So Naman, normally I think the channel is like to contact the chief statistician to ask the FAO, I think it's like FAO technical assistance on specific topics. So in this case, it's the implementation of the Yagri survey for you. And then I think, yes. In that case, you can address to chief statistician Petro Ghaneri, but make sure that also Joseph is the putter. We can copy to Joseph and FAO Bangladesh. Yes. This office is responsible for the indicators, no? Okay. Okay. And we can and we can send up my letter to also you. Are you want to put as a letter? I think, you know, it is better to copy you, you and then Narbha and then Margarita. Yeah, you put all the focal points in the city. Yeah. Yes. I would suggest to add to the email list, Sangita as well. She being a regional statistician. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Sangita should be in the loop. We are copying all our agenda, everything to Sangita. So she ultimately will be the responsible for the activities in the region. Okay. So can I show quickly what I have put together as a tentative? Can you see it? Yeah, you can see it. But you may make it bigger a little bit. Okay. Okay. So this is just taking from what we showed the first day as the objectives and outcomes of this meeting. So we said first one is a number of participants from initially it was only BBS, but now we have a participant from BBS, Ministry of Agriculture and Ministry of Food that were technically trained on the concepts and the methodology for identifying small-scale food producers and computing SDG 231 and 232. So that's the number one. Number two is underlying data issues, including how to go from the cashiness to clean the data, to organize the data, to aggregate at farm level, and to organize the database before making the compilation of the indicators. I think we went through all this and it looks like the participants understood very, very well the logic. And I think they are able to reproduce this with some maybe minor assistance from FFO. And this was done through presentations, discussions, practical exercises and demonstrations. So that's the number one. And then the second one is the available data in existing surveys in Bangladesh. And the requirement for the micro data that is needed to compute and report on 231 and 232 was identified and discussed. So this was our discussion of today. And then the gap, particularly in the last presentation by Mrs. Katoom, the existing data gap has been identified. And we just finished a discussion on the possible action plan in order to fill the gap with the FFO assistance. So that was the four main, three main results of the meeting. Now, regarding the next steps, here I am trying to put what we are doing with other countries. So since there are some available data in Bangladesh, particularly from the two sources, the Accensus and the household income expenditure survey, the first step would be to designate a focal point and a team in DBS that could work closely with the FFO team on using the, not using the capillary or over data, but using the data from Bangladesh in order to compute SDG 231 and 232. And I put here, this can be, should be done by August 2021, but I'm not sure this is realistic. Maybe we can at least start with the household income expenditure survey, which data is available because the census data, the processing is not completed yet, if I understood. And then to initiate the compilation of the SDG 231 and 232 with the existing micro data from household income expenditure survey and agriculture census by 15 September. And maybe by that time, the application that we discussed yesterday would be validated so that the big work would be to organize the data, etc. And once this is done, you just almost push a button and you get the indicators with the application. Or if you want to develop your own application, that's also fine. And then after that, the last part of our discussion now was to see how we can prepare a short action plan on how to improve and expand existing surveys to include the missing variables for computing SDG 231 and 232 with assistance from FOM. Maybe we should modify a little bit this statement in line with what Arbab has suggested so that to ask if VBS could write to FFO for assistance in improving the statistical system to be able to collect the data needed for the computation of the series of indicators, including 231 and 232. So this is very briefly what I propose as the conclusions and next steps. I don't know if you agree or you have any comment on it so that this is what we can take away from the training. Thank you, Mr. Laman, for your nice presentation and your our future action plan. But in your next step, you're going to initiate the compilation of SDG 231 and 232 with existing micro data from HAES 2016. And the aggregates are about the aggregates of census data is not available now, it will take some time. But our team, maybe if you guide our team can work with only HAES data, it is available, but aggregates of census data is not available. So it will be partial work maybe. Okay, I think it could be seen as a kind of exercise also. So real exercise so that the team of VBS go through all the process using real data from Bangladesh instead of fixtures of data so that when when the over data are available, they can just run the system themselves. And so we can correct that and put only the household survey. 16. There is a data gap in the household survey because there is no labor days and some related data, forestry data is not there. Yeah, but Naman, I think we can still start in the preparatory activities because we have, there's a questionnaire there for the census, understanding the questionnaire, trying to develop the program for we arrange the data, it really requires some time. Therefore, let's start with the preparatory activities. And once the data is ready, then it's just a matter of bringing the data and running the quotes. Okay, therefore, since we have at least a complete questionnaire, understanding the questionnaires and understanding the logic, how to prepare the data for the competition, it requires some time. Therefore, we can still start working on this, on this approach. But the actual data, yes, we can, we should be able to wait a little bit until the census is completed. So how about modifying the sentence to say, initialize the process for computing so that the process includes all the workers going from the questionnaire and looking at the first of all identifying the variables as Max has shown yesterday. So how do we go from the questionnaires, from the what variables to take and how we organize the database. And so when we get the data, we are ready for filling in the data and then computing the indicators. So only the process can be started at least for the census. How about that, Mr. Aladdin? And as to the process and after getting the complete data, then we can compute. Exactly. Just to maintain the, you know, the communication. I think we may have even a small sample data from the census as well, not the complete one, but just the sample data. So we can have with the sample data, we can start the work, you know. We don't need to have the complete data sense. So with the sample data, everything will be completed. Then when we have the complete data, we run the code, you know, the complete data. Okay. So with that, we will modify a little bit those two last sentences, the compilation and also the action plan thing to reflect what Mr. Aladdin just said and also what Arbab has proposed. So that with that, can we agree on this and go to the closing? Yes. Can I make just one point related to this issue of this algorithm approach? I think it is a good thing. We need to give it a try, but I still really wanted to re-emphasize, you know, relying on the existing system to be honest. I mean, this might be taken as a personal comment, but I really, really wanted to push to rely on the existing systems. The agrees thing, okay, we'll try. You know, I just looked at, you know, the Agri, the 50x2030 program, you know, what they have so far. They are already occupied with 24 countries and the program is covering up to 2024 looks like very tight. Therefore, you know, I mean, we should, I'm not saying that we should really push, you know, we should be able to find a way to come vis-à-vis the FIO management to include Bangladesh, but at the same time, we need to be to be aware of, you know, where we are in terms of 50x2030 as well. So, you know, I mean, I just wanted to give this thing as an input for future discussions. Okay. So I think we'll work together on to Jacob, Arbab, to see how we can formulate these last points of the action plan and to be as realistic as possible and see what the FIO really can do concretely to assist the country in dealing with this data gap. Okay. If there is no more comment, maybe it's time. Ms. Saleh, do you want to say something? Saleh? Yes. Yes, sir. I just want to tell you and only about the case data set. I think the simple census data set will take time. So, firstly, we can take initiative to calculate from HHS data. For the missing data, we can use as talk to you from any other health services, maybe from Bangladesh or maybe from the regional data set. So, if FIO can provide us a new file in startup, then we can go forward. Exactly. That's the same point that was. So, we can start the process and then as we go within the questionnaires and the missing data, we can find what is the statistical way of dealing with these problems and how to get the proxy to compute the indicators. Arbab, want to add something also? Yeah, just one point, Naman. And I totally agree with Yakub just mentioned. In terms of the 50 by 2030 team fully occupied in terms of them supporting countries. So, how about we subdivide the next steps into short, medium, and long-term steps? So, in the short term, we can have maybe some solution or some proposal. And then in the medium term, there could be a couple of proposals. And then for the long-term, I assume that agris and 50 by 2030 initiative will be more of a long-term solution. Maybe FIO will not be able to... And this is just my assumption. So, only the team can confirm the... But my assumption would be that they would not be able to support Bangladesh right away from the get go from the next year. Okay, so basically, concretely, we put the head... We put the two first points under the short-term activities. And then the last point would be put under the medium-term activities. And we need to reformulate it. Is that responding to you, Arbab, to your concern? Yes, yes, absolutely. Okay, so with that, I think we are a bit over the time, but it was for a good cause. Let's now go to the closing. So, if Dr. Alauddin and Dr. Amroul, Alauddin for BBS and Amroul for FAU, want to say a few words? Let me say first, and then Mr. Alauddin may conclude from what BBS said. Just to know a few words, we are already... We've crossed the time limit. So, with a few words, I just want to thank, you know, you, thank the headquarter and your team, Ida and other resources persons, those who are involved. And it was the fruitful training, actually. And so far, all of the trainings they know on 2.1.1, 2.1.2, even 2.4.1, 5.8.9, all are interactive and very much successful. And BBS learned a lot, their capacity just enhanced a lot, but it is not turned into action immediately. So, this learning may fade out one time, actually. That is my worry. Otherwise, you know, we are successful. And this program that was taken from FAU headquarter, supporting Bangladesh, it is so far successful. Apart from this COVID impact, otherwise, this training could be face-to-face. The learning would be much more efficient. But still, I think, you know, the training is very much successful. From my side, I am satisfied because, you know, BBS is also participating in the sports and practical sessions. So, there is some skill developed that will be, you know, a permanent asset for BBS. So, with this word, I want to thank BBS and also line ministries, the ministry of food, ministry of agriculture, they have nominated some participants. I hope they have also learned something from this one. And do not present our FAU Bangladesh team. I also want to thank FAU Bangladesh part, because I am a temporary consultant for them. But FAU Bangladesh is a long-standing friend of BBS. Whatever BBS is requesting, they are trying their best to pursue the headquarter and then taking some support. And with this one, I also want to thank our TV station, because, you know, this was a long, you know, discussion with him, because he thought initially that enough training has been given to Bangladesh, why more? It was his concern. But, you know, I and BBS also explained, you know, all of the training is not dedicated to BBS. There are several trainings given to other ministries, other sections of the, you know, ministries, but BBS did not get all of the trainings. So, this time only, very much dedicated trainings for BBS. So, BBS learned a lot and capacity as they have learned. With just a few words, I will request BBS to keep contact with all of the local points. Now, I think, you know, the communication is very much easier. Whenever you need any support, any suggestions from headquarter, please write to them. Or even if you want to tag me in the discussion, I am always there, even if I am not with FAU in future. So, I will be always there with you. With these few words, I want to thank all of you for this successful training, not only today's training, these two indicators, also as Arbab is here, and through this session, I want to thank all of the local points for all of the six indicators. Initially, five were in plan, but we could handle six, this is kind of an extra for us, you know, total six. So, thank you very much, you know, for giving such such effort from your side. And thank you, BBS and other players. Thank you, Amirul and Mr. Aloudin. You want to also say some words. Thank you, Professor Amirul. And I also thank Mr. Naaman, Ms. Aida, and Mr. Faridun and your all team, two, three, one, one and three, one team, to giving us this training and our participants learn a lot from you. And I also thanks to others, team, just one team representative, Mr. Arbab Khan is here, 2.4.1 and team 2.1.1 and 2.1.2. And five day one, I also thanks to all in favor of from BBS and also to FPO Headquarter and FPO Bangladesh to arrange such a wonderful training for us and our participants learn a lot. And I just, it is the initiative and we can move forward in future. We will communicate each other. And I hope we will get support from you in every time. And so that we can compile the SDG indicators, we can support the data for SDG indicators. And I also again thanks to Professor Amirul to organize this in such a nice way. And he is always a friend of BBS. And he is always with BBS. And I hope all of you with BBS for calculating these indicators, collecting data and compiling the indicators and hope our communication will continue. So thank you again and thank you all. Yeah, thank you, Dr. Alauddin and Mr. Alauddin. And you are right. I think Dr. Amirul has before we close, you know, let me especially thank Mr. Yakub because, you know, sometimes I had, I had to go to him, please, you know, feed up the process from the headquarter because sometimes, you know, I was waiting for response because all of the team at the headquarter are very busy. So time wise, we are also like in the end, our program will finish by November. So we have to finish all of the trainings. I sometimes, you know, requested, you know, Mr. Yakub, he's a longstanding, you know, friend of BBS, I think, you know, through the MS project. Yeah. So always whenever I find any problem, I always like to write to him. I saw that you copy always to Yakub and Dr. Amirul, you have been very, very instrumental to this because I see you have been really pushing this training. Once we had some delays, etc., you come back. So that was very, very, very useful for organizing this training. And I just want to thank everybody also, all the participants, to thank also the BBS, the Minister of Agriculture, the Minister of Food, and particularly the FAO office also for this organizing this training. And it has been a very intensive and interactive sessions because it was not just one way lecture that it was really interactive. And this is what we wanted from the beginning. So I hope by now the concepts, methods, and data issues are much clearer. And I'm sure because the participants went through the process. And finally, we agreed on kind of next steps in order to keep the momentum. So we don't want to lose the momentum. So we want to, after this training, not to everything to should die, but to continue the collaboration and the momentum for this, what we had capitalized during this training. So thank you all. And now we move to the last part of this training, which is the group photo. And I give the floor to Faridun to tell us how we should do in order to take this. Thank you, Naman. As you will see, I sent to everyone the link with the passcode. You need to go out from this webinar and log in with the link which I sent you. It was sent in three o'clock of Bangladesh time, all around this time. So for now, we should go out from this link. So we go out and we click on the link and put the passcode and then we take it from there. Yes, the passcode is from 1 to 7, 8. Okay, let us leave and rejoin. Yes.