 Okay, this is the second in our series of videos about MOSMOD, and in this video we will be introducing policies, and in particular discussing the definitional policies, and introducing the tax and benefit policies. First, I want to go back to the user interface, which we looked at in the last session, and look at that in a little bit more detail. As I've already indicated, the policies are listed in the second column in the user interface. That's their actual names. The policies are usually named for the output variable. They're clearly described in the comment column, and classified between definitional policies that begin with D E F, tax policies that begin with T A X, social insurance policies that begin with S I C, and benefit policies that begin with B E N. Now the ordering of the policies is important, and that ordering is called the spine, and then you can see there's numbers running from 1 through to 18, and for each of the policies, and that's the order in which they're simulated. The order sometimes doesn't make any difference, but for some policies you need to have the result of a previous simulation before you can simulate the policy you're looking at. So the order is crucial and needs thinking about. So for example, if a policy to give some kind of social benefit required that the person wasn't receiving some other kind of social benefit, then it would have to be placed after the other social benefit, so that the simulation could make a check to see whether it had already allocated the previous benefit, in which case that might disqualify you from the subsequent benefit. I hope that's clear, but basically ordering is important, and that ordering list is referred to as the spine. As I indicated, there are three types of policy, there are definitional policies such as the first policy, up rating. There are tax policies both direct like income tax and simplified tax, but there are also indirect taxes like VAT, and then there are benefits in which case mainly the direct social support programme, and I also indicated that there are social insurance policies too. I'm going to go through the definitional policies in more detail, and when I do so I will actually flip to the model and talk briefly about each one of them. So there are up rating factors, that's important because the data set is a data set for 2008-9, yet we've got policies for 2015, so data from the data set needs to be up rated in various ways. There are income policies, income concepts, excuse me, those are defined in income lists and we'll go through those in a bit more detail. There's a definitional policy on tax units, and tax units describe the composition of households and if there are special definitions of other groupings within the household they're contained in these tax units, so a couple may be defined in a tax unit, a family as a subset of a household may be defined in a tax unit, etc. Constants are, as the word indicates, things that don't change within the model, it's where values of benefits are often put, it's where tax bans and tax rates may be put, it makes it much easier to update the model if you've got those figures as constants rather than hard coded into the model. And then there are two expenditure variable policies relating respectively to value added tax expenditure and excise duty expenditure. And then there's a standard output at individual level definitional policy. Now one and four, that's up rating factors and constants definitions will need probably annual up rating, I mean certainly the up rating will and often the constants will. Two and six may need amendment if a new benefit or tax is introduced, so that's the income lists and the output policy. The definitional tax unit will rarely need amending, basically they're set to reflect arrangements in the particular country. And of course the expenditure variables will only need amendment with the introduction of a new data set. Now I'm going to take those one at a time and talk in a little bit more detail. So if we look at up rating factors, this policy inflates values in the input data set to bring them up to date and is used with the up rating indices country tool. So as I said the underpinning data set for Mods and Mods is currently 2009 and therefore for 2015 all values need to be adjusted for inflation. What inflator is the CPI, that's what will be used for things that aren't specifically specified but other specific inflators can be specified for particular values. For example, earnings are inflated by wage inflator calculated by looking at annual changes in the minimum wage averaged across all sectors. Now let's go and look at the model and look at up rating factors. So here's the model and if we look at the model the first policy is up rating and let's look at it here. So the data set is specified, this is the data set we're working on and then the default factor and then a special factor for wage and self employment income. Now they're referred to as a factor that begins with the dollar sign, a bit like a constant as we'll see. Now they're actually in the country tools there's an up rating indices section which is where you define these factors and the factors are defined here. You can see that there's a CPI total, CPI food, non food, general wage inflator, CPI for alcohol, tobacco and fuel. These are differently used in different parts of the model but they're all contained together in an up rating indices section and additional years can be added, years can be deleted and that manages if you like up rating. So that's where those values for these three factors that are in our up rate policy are stored. I'll close that and go back to the model, sorry the presentation. Okay the next is I've mentioned constants and the constants contained in MOSMOD are shown actually in this screenshot. They mainly actually relate at the moment to the rates for excise duty, VAT etc. Now that doesn't mean they're the only constants that exist within the model, there are other constants and one of the tasks on this year's agenda is to move more of the hard coded amounts within the model into constants. Now I've mentioned income concepts or income lists, there are many different income concepts and required to simulate taxes and benefits. So for example the income components included in the means test for social transfers or the income components that make up taxable income. So the policy Ildef underscore MZ specifies which income components relate to each income concept or income list. It's not necessary often to amend this particular policy as I indicated earlier but it's something that one needs to be aware of. So here are the income lists in MOSMOD itself and you can see that it covers taxable income used for non-employment income, taxable income used for PAYE, for simplified tax etc. I'll move to the model to actually show you these in more detail. So if we look at the income lists, first of all income list, taxable income used in personal income tax for 2015 for non-employment income if no simplified tax is payable. When we look through the model in the final video I will explain that in more detail but important to know that the income list is slightly different in cases where there's no simplified tax, basically it includes self-employment income. So here is the income list, you've got income from employment which isn't relevant because that's dealt with directly in the model. There's other income, income from property, income from agriculture, income from self-employment income from interest. Now what an income list does is basically adds, if there's a plus sign, those incomes together and puts them in that income list which is a bit like a composite variable. It's the way to think of it I think. For some income lists, if you put a minus as opposed to a plus then it will subtract that from the total. So there are income lists for simplified tax where you have self-employment turnover added to agricultural turnover for the simplified tax income list for example. And then some income lists are actually standard income lists that are defined for example, original income and disposable income etc. Again these familiarity with these will come with familiarity with the model and in my final video I will go through all of the income lists. Okay, now back to the presentation, there's now we want to talk about tax units. Tax units are groups of people, I sort of indicated that, individual, couple, family, household. Family and household may be synonymous in Mozambique and some countries they're quite closely defined and don't include necessarily all members of a household. And tax units determine who's included during the policy simulation for example. Local income tax is calculated now at the individual level. One of the means tests for the direct social support program depends on per capita household income being below a certain amount. So that's calculated at the household level and these levels are specified and have to be specified in each policy but not usually necessary to order these definitions. So I'll just quickly again shift to the model itself and give you a site of the so-called tax unit definition. So actually in MozMod we have a household, as you'd expect that's all members of the household. We don't have a family definition in MozMod in the conventional way but we do have a particular kind of family that we use in some policies and we have obviously individual but we also have couple. So these are one, two, three, four different tax units that are defined. And the tax units we'll say something about and I'll give you an example, sometimes quite complicated to grasp but this particular family unit includes, it's a subgroup of household, it includes dependent relatives, it has a condition for partners and a condition for children. Back to the presentation, all of this will become clearer as one uses the model but as I said these particular concepts are not likely to need amendments very often. So we then move to tax policies. We have a policy for income tax, simplified tax, value added tax and excise duties. So two direct tax policies, income tax and simplified tax and two indirect value added tax and excise duties and these can be extremely complex and indeed income taxes is very complex in Mozambique, simplified tax on the other hand is very straightforward and they're only likely to need amendment if for example we're testing out new ways of financing new benefits or if the tax system is radically overhauled. I'll go just quickly to indicate tax policies. Let me start with the straightforward turnover tax. The turnover tax is a tax for people in self employment whose turnover is less than 2.5 million Metacals a year and it's a very straightforward and simple tax. The person, often an informal trader or whatever, doesn't have to keep accounts, they're simply charged 3% on their actual turnover, it's a very straightforward policy. Income tax on the other hand is a much more complex policy. This is the PAYE income tax policy and as you can see if I fully expand it, it actually starts here, there's actually quite a lot of computations which conclude down here. No need to get baffled by this at this stage, again with experience of the model all of these and method, what do I want to say, clear thinking these policies can be perfectly well understood by anyone that can understand simple formulas, they really are basically simple conditions of entitlement and simple formulae so really nothing to get concerned about and probably will require quite infrequent amendment. Okay moving on to benefit policies, we've got really two main ones, the direct social support program and the basic social support program and we're going to cover these in a future session so I won't mention more at this stage. I talked about the spine and it's worth repeating, it's a list of policies simulated by Mozmod, each policy is numbered and ordered in the main window and policies are executed in the order by the program itself so it's imperative for example that policies which depend on information generated by an earlier policy are positioned after that policy and that's what I was indicating earlier so I think the activity associated with this particular presentation is to explore the model, identifying different types of policy and becoming familiar with the definitional policies which form the basic framework of the model and then in the next session we will look at social benefit policies in more detail and indeed in that session we will look at the construction of each policy in terms of functions and parameters and that's all for this session, thank you.