 Okay, so this is the joint work with our colleagues, Quang Nguyen and Zero Calito. And just before the conference, we got the good news that it was accepted, you know, with the journal, World Development, so of course you have on the full version coming out there soon. If you want to look at more details. So for this paper, we are looking at the impacts of the COVID-19 pandemic on employment outcomes in Vietnam. And we looked at, you know, earlier, you know, this paper has been, we have been doing it for some time. And we did it, you know, when we got the data for the labor force survey data for Vietnam in 2020, you know, with like a span, you know, like the pandemic. And basically, I can go quickly. So our contribution is that, you know, we contribute to a growing small, right, but growing literatures on the impacts of the pandemic on poorer countries. So even though we have a lot of study, right, for richer countries now, but for poorer countries that use nationally representative survey data, we still don't have a lot of evidence. So here, it's very interesting that in this panel, right, we see more evidence, right, coming up. But you know, but if you look at the literature, you know, most of the public studies, I mean, so far, only use, mostly use high-frequency phone surveys. And there are multiple, you know, there can be multiple problems. So with high-frequency phone surveys, if we don't use the data well, right, adjacent problem, under coverage problem, selection bias, and so on, right? Yes. Okay, so in our, the second point, the second new contribution with our paper is that we analyze annual and, you know, very large-scale labor force survey data for Vietnam. So which allows us, you know, to break down, you know, the impacts, the heterogeneity impacts for different population subgroups, for example, by educational level, by gender, and so on. And on top of that, we can also look at, we also did look at a wide range of labor indicators, for example, unemployment, temporary layoffs, you know, labor force participation, labor income, right, because it's labor force survey, so it's like labor income. And then we also focus, we also pay a lot of attention to low-income workers, you know, the proportion of the low-income workers who were affected, you know, by the pandemic and, you know, resulting impacts on wage inequality, you know, for the whole countries. So very briefly about the background for Vietnam. So Vietnam, you know, it goes like a lower middle-income country, but, you know, like it puts up, you know, quite a good fight, you know, against the pandemic, at least earlier, you know, like for earlier rails. And, you know, because here we, our data, right, until the end of 2020, so, you know, and the lockdown, you know, the pandemic lockdown happened in early 2020 in April for about two weeks. And so, for the survey, right, and we can look at like several years, you know, before the lockdown and, you know, and about, you know, several months, actually, like about two, three quarters, you know, nine months, you know, after the lockdown. And so the good thing about Vietnam is that the government, you know, they put up a very strong fight against the pandemic and so they apply, you know, very rigorous measure right at the beginning of the pandemic. So if you look at the efforts, check the surveys, right, then you can see that the index, you know, the government response index for Vietnam, you know, was very, very strong, you know, right at the beginning of the pandemic, early in 2020, partly because the country, you know, like has a safe border with China, so maybe like more information, you know, like is available at that point, yes. So by the end of 2020, you know, there was only 78 deaths, you know, our population of only 100 million. So that's quite good, you know, if you compare with other countries, like the U.S, right, the UK or France, Italy, right, we see a lot of deaths. Of course, I mean, this number, you know, went up significantly, you know, by the end of 2021. But still, you know, if you compare to the population, you know, still, there is still quite a low number. And by the end of 2020, you know, that's a very, very low number, yes. So given this background, right, so we wonder, you know, like whether the pandemics and in particular the lockdown, the pandemic lockdowns have any negative impacts on labor outcomes, on the outcomes that I mentioned, right? And whether there is some kind of heterogeneity of impact for different population subgroups and which occupant sector, you know, are more, you know, impacted and furthermore, you know, we pay more attention to lower income workers. I mean, so, you know, that are more likely, right, to be affected in any, you know, major economic crisis. Okay, so, yes, I already talked about the data. So, and basically, you know, for the equation, our main equation is like a different in different approach, where basically we interact, you know, like COVID year, remember that we have a five year, no six years of data, right? Two and a fifth in 2020. So basically we have a several years before the COVID year, quickly 2020. And then we interacted that COVID year with the Kutu dummy. So basically here we have Kutu two and then we interacted with COVID year. So the coefficient of interest are beta five, beta six and beta seven, you know, basically just different in different, you know, like a coefficients. So this model, you know, we only give us, you know, the overall impacts of the pandemic, right? Because deep, deep in deep, so it can pay before and after. But if we want to look more closely, right? Into the local impacts of the lockdown, right? If we want to identify the local impact of the lockdown, then we need to augment it, you know, with RDD, you know, with regression discontinuity design, where we identify, you know, the cutoff point where the government implemented the lockdown in early in April, you know, 2020. And then on top of that, you know, we also supplement it with a combination of RDD and deep in deep and RDD and cheaper, different, yes. So you know, so that, you know, to make the model more rigorous. Why? I will explain more later. Okay, so basically here, so this is a RDD model where basically we, as I mentioned, we set the month equal to zero for April, 2020. And then, you know, before that, you know, on the month have a negative value. After that on the lockdown months have a positive values. And then as a local impact, the local impact is measured by delta one. We give you like a standard, you know, sub RDD model. But again, you know, we want to control for, because here like we have monthly data, right? So we want to control, you know, whether there is some seasonal effects, you know, with employment. So in that case, we can also interact, you know, like a lockdown month, you know, with the COVID year. And then we add a bunch of control model. Again, it is pretty standard. Well, not very commonly used, but pretty standard, you know, RDD and combined with RDD, yes. And then on top of that, you know, because there is some concern, you know, from some reviewers that, you know, ZIT model is not rigorous enough. And as I complained that, we need to bring in some kind of geographical location, right? Some kind of zero for variation. Because here, if you remember on here, like a Thai variation, right? Lockdown month, multiplied by COVID year. So, you know, the question is that, you know, if there is some kind of, is there any kind of zero for variation that we can bring in? For example, to show that different locality, different provinces in Vietnam, you know, have different ways, I mean, to fight the pandemic, or they may have a different, you know, the measures, you know, maybe, the rigorous, you know, to the different, you know, extents, right? So, to address that concern, then basically we bring in the RDD plus cheaper, different model, where basically, you know, we interact everything. And the coefficient of interest is N-5-2 here, where basically, you know, we multiply COVID year, quarter, two and four, and then a province in central Vietnam, Da Nang, we actually implemented two ways of lockdown in June 20. That's the only province in Vietnam that implemented two ways of lockdown. And so, so that RDD, so this last model is in comparison to the other provinces in the country, yes. So basically, that's our model. And very quickly, I'm going to show you the main reasons. So the main reasons for the deep and deep model, sorry about this, it's like some problem in Excel. But basically, if you look on the deep and deep model, right, and you see that, of course, they have the expected size, where basically the pandemic, it, you know, it increases, right? The unemployment rate, and then, you know, it also increases the temporary layup rate, and then it, you know, have negative impact on all the other outcomes, variables, right? For example, having a wage job, having a job with a contract, like a former job, right? And then, you know, like a, but number of working hours, you know, the pandemic doesn't appear to have any impact on the number of working hours. But then it, of course, it has a negative impact on monthly wage, and then it also increases, you know, the proportion of low wage workers, those who work below the minimum wage. So, yeah, so we do see that for all the four different cultures, even though we see some, like some declining, you know, effects, you know, like to work the end of the year. I'm gonna show you more, I'll wrap it later on. But again, you know, only thing we can tune for on the fixed effect, right? The check feature effect, proven fixed effect, here fixed effect, and on the other variables that I mentioned, yes. Okay, so now, so now coming to the RDD and deep and deep model, right? We, we, we want to look at the local impacts of the lockdowns, right? Because now we want to pinpoint the mechanism. The reason before is like the overall impact of the pandemic, you know, could come, you know, from various factors, right? So now people can say that, okay, what about some seasonal factor, right? Because maybe employment change in them, it can happen, you know, like for different types of year, right? So here in response to that, we have this model and basically, you know, we have a very qualitatively, we have very similar reasons. Of course, the magnitude of the impacts, you know, are different, right? But we have see very, very similar reasons. Where basically the lockdown, you know, have a positive impact on unemployment rate, temporary labor rate, but then it have a negative impact, you know, like on the, having wage job, having former job or, you know, monthly wage income. And finally, you know, we also saw the RDD come by with cheaper difference. And again, we qualitatively, you know, we see the similar reason. And here we just saw like a different low between tax, you know, where we play around, you know, with different sample, right? For example, for the control sample, we can either go on the way from 2015 to 2020, or we just go from 2019, only 2019, or we go from June 17th, 18th and 19th, right? So like a different, you know, like a control group here, you know, to make sure that, you know, the reason are robust. Okay, so here is like a more detailed analysis where basically we saw the impact of the pandemic by the month, right? So here, for like, I just saw it for the following outcomes. And you can see that here, the impact, you know, of the pandemic, you know, appeared to be, you know, strongest earlier in the pandemic around April and May 2020. And after that, it tapers a bit and, you know, remain a bit stable, especially for temporary layout, you know, the impact is very strong, you know, early in April 2020. But after that, you know, it's almost like go to no impact near the end of the year. But except that, you know, for low income worker, you know, they still seem to be affected more to this year, but later on in the year, they are also very much affected for which, you know, we are analyzing, you know, like more recent data, you know, to find out exactly what is the reasons. Okay, so as I mentioned, we play around with a lot of broadband checks. If you are interested, please refer to the paper. Basically, we run on type fixed effect with a fixed effect with control variable, without for control variables. We play around with the different control gears. And then we also run some placebo test where, you know, we pretend that, you know, the year of the treatment is not 220, but like 2019, 2018, for example, right? So that type of placebo test, you know, to make sure that, you know, the reason robust. Okay, so finally, two more graphs. One here is that I saw some heterogeneity of impact where basically you see that we do see some heterogeneity of impacts. But very surprisingly, for Zou, Zou with like more educated people who are workers, who are more educated, you know, like are more affected, you know, than Zou, you know, like her lower education, right? And also we see that, you know, for younger workers, they are more affected, and older workers like Zou who are close to the retirement age, then they are less affected, you know, which makes sense, right? Because then they have a steady, you know, like a pension or wage income or something for that. We don't see much difference between men and women in Vietnam. They are affected, you know, more or less similarly. Okay, but now maybe the last, maybe interesting thing is that here we look at the proportion of low income worker, right? Zou in the poor consumption quintile. And Zou, you know, working below the minimum wage. And here we can see that, you know, they are more affected, you know, the bottom 10% or bottom 40%, yeah. So basically, you know, lower income workers, wage worker in Vietnam, you know, are more affected by the pandemic than other groups of workers. And maybe in response to a point about why inequality. So maybe I'll respond here in this case is that, you know, we see more income reduction for lower income worker, you know, than higher income worker. So that's why, you know, overall, right? We see like a more gap between the two groups. So that's why we see, you know, like a more wage inequality. So in the paper we did conclude the table, you know, where we saw different inequality indexes, like the genie, Thai index and so on and so on. And we do see, you know, at least some temporary increase, you know, in wage inequality for only groups. Okay, so I can stop here. And because we run out of time, yes.