 My presentation is entitled, Assessing the impact of the recording VAT intervention in Zambia. It is like the other studies, a joint collaborative study between the Zambia Revenue Authority and the UNWIDER. In terms of, well, what I'll talk about, I'll give a brief background, then talk about the method and the data that we use, the objectives, as well as the key results and policy insights. In terms of the background, well, Zambia, like many other developing countries, have large informal sectors. There's also these businesses or these sectors are categorized by poor bookkeeping and they usually don't have systems in place, systems and controls to record the business transactions. And in general, these countries are characterized by limited ICT systems and interconnectivity of systems. There's also generally limited expertise to uncover sophisticated tax avoidance schemes. Hence, recently, withholding mechanisms have become quite popular in third world countries to help address the low compliance that arises from the structural issues in the economies. I also want to state that in Zambia, VAT is one of the major tax types, I think contributing an average of 25% in the last five years. We have seen a gradual increase in nominal terms in VAT collected over the years, over the past decade or so. However, the countries still continue to face challenges of low return filing, limited self-enforcement mechanism in the VAT, which is supposed to be inherent in the design of the tax type. Also, under reporting and misreporting, which was then resulting into a narrowing gap between what we were receiving as the formal claims through the return filing process as against the gross collections that the country was recording. And to help address these anomalies that we were observing, we decided to introduce the withholding VAT mechanism in 2017. So the way the mechanism works is that certain players in the economy, usually these are large, very large tax payers in the top 10% of the contribution to gross revenues if you want or contribution to gross domestic product are selected as the withholding agents, including the government and its alive institutions. Then these entities, whenever they deal with suppliers for valuable sales, they withhold the VAT instead of making the payment together with the VAT. So currently, the withholding rate is at 100% of VAT. The standard VAT percentage in Zambia is 16%. Therefore, withholding agents currently withhold 16% of the VAT as they are paying those they interact with. And so far, we have slightly over 100 VAT withholding agents which comprised of mining companies, the large mining companies, commercial banks, as well as the government and its allied institutions like para-staters. I should state that currently in Zambia, over time, we've had an average VAT registered businesses of about 18,000 out of which the active tax payer population is now at around 7,000, but before that it was much lower before the withholding mechanism is patient. This active tax payer population for VAT was much lower than the average we're observing now, which is at 7,000. So in terms of what the study was trying to uncover was just to answer the questions which is what is the impact of the withholding mechanism in Zambia after its introduction. Then what is its impact specifically on reported VAT revenues as well as what is this effect on the behavior of these registered entities, especially when we check the behavior of these firms according to location or the economic sector in which they're operating. So to effectively answer these questions, the study employed the different methodology. The treatment group was made up of firms whose VAT had been withheld before, while the control group was those firms who have never had their VAT withheld. This was largely concentrated on the small and medium-sized firms because we know that the large firms, like I explained, were already in that category where they were largely formalized and hence the problem of the disappearing VAT was not so common amongst the large firms. So the study period is 2014 to 2020 and we utilized a monthly VAT returns, detailed monthly VAT returns data obtained from the Zambia Invenient Authority systems. The authority, I should mention, digitalized its tax collection system in 2013 and since then the taxpayers were required to submit detailed transactions up to the invoice number level in their VAT returns. Hence, there's this rich micro-level data that is available for such a study. In terms of the observations that we're dealing with, the total VAT returns, which are received on the monthly basis for the period 2014 to 2020, were about 13,000, amounted to 13,309 firms, whereas the firm level withholdings, this is now after introduction of the reform in 2017 to 2020, we had 4,559 unique firms that belonged to the Withheld VAT. So in terms of the treatment sample, we're looking at suppliers who had interacted with holding agents consisting of about 4,488 firms and the control group had 9,200 firms. In total, we were dealing with about 53,000 observations. In terms of just a brief on the descriptives, I think the majority of the filings were from the wholesale and retail sector, which accounted for 43%. And then we saw a spike in the count of returns in 2017 and 2018, largely owing to another intervention that ZDRI carried out during that period, which was the tax amnesty, where taxpayers were allowed to file their old returns and their penalties and interest were... In terms of the findings from the diff and diff, if you just look at the descriptives, checking whether the parallel trained assumption was holding, we see that indeed it was holding for the variables starting with the total sales, the total output VAT, as you can see from the red dots indicating the treatment group was indicating the control. Basically, we looked at four main outcomes here because we observed that these were the outcomes that the withholding agents looked at when they interacted with the supplier. I would show the impacts here. We realized that there's a significant positive effect on these outcomes that we looked at. On average, you would see that the reform had about a... let's say about 10% increase for these outcomes, which are value-added total sales, and then output VAT. We observed a greater increase in output VAT. The difficulty we had here was with reported purchases, which is quite ambiguous. We attribute this to the fact that reported purchases by the supplier is not directly observed by the withholding agent. You would see that there is a bit of leeway for the supplier to misreport purchases because it's not cross-checked by the withholding agent. All in all, in three out of our four outputs, we observed that there's consistently increased post-reform. We also go ahead to do some heterogeneity analysis in respect of the location of the firms and then the industry where the suppliers are found. We realized that in those regions where you have a concentration of suppliers like the copper belts and then the lewisaka, you would see that we observe significant increases in value-added and output VAT, although the impact on sales and purchases is not significant. Those are the effects in terms of the location. In terms of we do also effects on the industry, but we don't show this here now. For policy, you'd see that ideally we can conclude that the withholding mechanism improves the compliance behavior when there is a shift in the reporting structure where now it's not the supplier who's reporting, but now it's the withholding agent that's reporting the tax to the administrative authority. Basically, the design accounts for everything. What we are thinking about is that it should take into account administrative and compliance costs because we don't calculate that here. This is one thing that we should look at. Then also we observe that we should try to localize the incentives here where selected regions should have specific withholding agents rather than the broad sector withholding agents that we have now in the ZRI. I would stop here and sorry about the bloop. Thank you.