00:03 Uh,
00:03 welcome everybody.
00:05 Um,
00:06 this is our May edition of the policy research talk series.
00:11 As many of you know,
00:12 these talks give us the opportunity to present
00:14 work coming out of the research department,
00:16 um,
00:17 and in this case with some of our collaborators.
00:20 Um,
00:21 with the goal of sharing the findings within our own department,
00:25 across departments within the World Bank,
00:27 as well as with,
00:28 uh,
00:28 others outside of,
00:29 uh,
00:29 the World Bank itself,
00:31 I'd like to welcome our online audience on Webex
00:35 and the live stream as well as our in-person audience.
00:38 Uh,
00:38 thank you for coming.
00:40 Um,
00:41 today I'm really pleased to introduce our speakers,
00:43 Oyebola Okonogbe,
00:45 Pierre Bakas,
00:46 who are two researchers in the Development Research Group,
00:50 along with Anne Brockmeyer,
00:51 who's with the macroeconomics Trade and Investment team of
00:54 the Equitable Growth Finance and Institutions Vice Presidency.
00:58 I promised myself I wouldn't use any acronyms today,
01:01 so
01:02 I'm spelling everything out.
01:04 Uh,
01:04 our 3 speakers will be presenting a work
01:07 related to the topic of taxation for development,
01:10 and,
01:10 uh,
01:11 for those who don't know,
01:11 actually all 3 papers were featured in a recent
01:15 issue of the Journal of Economic Perspectives,
01:18 uh,
01:18 so we're getting a personal,
01:20 uh,
01:21 presentation of some very high profile work.
01:24 Um,
01:25 Oyebola is an economist in the human development team of the
01:29 research department whose research focuses on governance and political economy,
01:33 including policies related to public finance,
01:35 nation building,
01:36 education,
01:37 employment,
01:37 and gender.
01:39 Pierre is an economist,
01:41 uh,
01:41 on the macroeconomics and growth team
01:44 whose research focuses on public finance in developing countries,
01:47 in particular,
01:48 optimal tax design and challenges to tax collection,
01:52 uh,
01:52 resulting from tax evasion,
01:53 informality and differences in economic structure.
01:57 Anne Brockmeyer is a senior economist and the global
02:00 lead for tax data analytics at the World Bank.
02:03 Her research lies at the intersection
02:05 of public economics and development economics
02:08 with a focus on tax system design.
02:12 I'm also really happy to present to to welcome today our discussant
02:16 is Manuela Francisco.
02:18 Uh,
02:18 Manuela is the global director for the macroeconomics
02:22 Trade and Investment Global Practice in the
02:25 Equitable Growth Finance and Institutions Practice Group,
02:28 soon to be renamed Prosperity.
02:31 It'll actually be less of a mouthful.
02:33 Uh,
02:33 it's a position she started on the 1st of January 2023.
02:37 Uh,
02:38 Manuela joined the World Bank in 2005 as
02:40 a country economist in the Africa department.
02:43 She's since held various positions,
02:45 including a director of credit risk in the Chief Risk Officer vice presidency,
02:50 and before that she was a practice manager,
02:52 um,
02:53 for MTI,
02:54 and I already spelled that out previously,
02:56 uh,
02:57 in the South Asia region.
02:59 Uh,
02:59 before joining the bank,
03:00 uh,
03:01 Manuela was an assistant professor at the Universidade de Mino
03:06 in Portugal.
03:07 She holds a PhD in Economics,
03:09 International Macroeconomics from the University of Nottingham.
03:13 Um,
03:13 I'll start by asking Oyebolo,
03:15 Pierre,
03:15 and Anne to talk for about
03:17 collectively 45 minutes,
03:19 and then we'll hear from Manuela for,
03:21 uh,
03:21 for,
03:22 uh,
03:22 10 minutes or so,
03:24 and we'll conclude the session with a Q&A
03:26 with the audience both in person and online.
03:28 If you have a question online,
03:29 please use the raised hand feature.
03:31 Um,
03:32 uh,
03:33 or just signal in the chat that you have a question,
03:35 and I'll call on you.
03:37 Uh,
03:38 just a reminder,
03:38 we are recording the online version of this
03:42 and also if you're not speaking online,
03:44 please do mute your microphone.
03:46 With that,
03:47 over to you presenters.
03:50 Thank you very much Dion.
03:52 And thank you all for being here it's such a pleasure to have this room
03:56 filled with everyone and also for everyone that we have joining us online.
04:01 Today I'm going to be talking on a subject that is of great interest
04:06 to policy makers to
04:09 international organizations like the World Bank and to researchers alike,
04:12 and this is the question how can lower income countries
04:15 collect more taxes?
04:16 I'll be focusing on the role of technology.
04:20 Tax agents and politics,
04:21 there's a little bit of a.
04:25 Yeah,
04:26 if you're online,
04:26 please can you make sure that you're muted?
04:28 Um,
04:29 I see there's one person who's not muted.
04:32 Also,
04:32 we'll pull up in the room.
04:35 Sorry,
04:35 it was my hair.
04:39 OK.
04:40 Um,
04:41 so
04:42 this is joint with Gabe
04:44 Trek who's University of Pittsburgh.
04:47 So the starting point for the talk today is that
04:50 we all know that taxes are important for development,
04:53 so governments need tax revenues to invest in public goods,
04:56 to provide public infrastructure,
04:58 to invest in human capital,
04:59 to provide social protection.
05:01 The needs are endless.
05:02 At the same time we know that there are
05:06 at the moment,
05:07 um,
05:07 low income countries tend to not collect enough taxes.
05:10 So
05:11 low income countries collect about.
05:14 10-12% of GDP
05:18 in taxation revenues on average
05:20 compared to about 30%
05:22 in higher income countries
05:24 and.
05:26 One thing to note is that this is.
05:30 Very similar
05:31 position to where
05:33 today's high income countries were a century ago.
05:36 So on the right map we have
05:39 18 high income countries
05:42 and their progression over time
05:44 and we find this.
05:47 Relationship which suggests that perhaps tax capacity
05:50 is a byproduct of economic growth.
05:54 We examine the data for this,
05:56 and we do not find evidence to support this.
06:00 In fact,
06:00 what we find is that if you look on the x axis,
06:02 this is the change in the economic growth,
06:05 change in GDP per capita over 30 years from 1990 to about
06:09 2020.
06:10 And on the x axis is the change in tax to GDP of countries.
06:13 This relationship is,
06:14 is pretty much flat.
06:16 And all this suggests that
06:18 countries actually do need to make this intentional investment in their tax
06:21 capacity if they want to increase their level of tax collection.
06:27 Another thing to note is that the tax
06:29 rates and policies that lower income countries have
06:33 on paper are very similar to what higher income countries have
06:36 and so the fact that we see these gaps
06:38 in collection suggests that there is an important role
06:41 for administration,
06:42 meaning the ability to actually implement the policies that are on the books.
06:46 And today we're going to be discussing three dimensions
06:49 of tax capacity that governments need to invest in
06:52 if they want to increase tax collection.
06:54 The first one is what I'm referring to as identification capacity,
06:58 and this is meaning that governments need to be
07:00 able to identify the tax base and taxable entities,
07:04 and this is by no means trivial because
07:07 if you take a place like Africa for example,
07:09 where 45% of people do not have a government issued ID,
07:12 they're not known to the state.
07:14 If the state in that kind of setting wants to collect taxes,
07:17 the first point of call is actually
07:19 increasing your knowledge of who your tax base is.
07:22 Next we have detection capacity,
07:25 which is the ability to verify the amounts that are owed.
07:27 What is the true tax liability,
07:29 and many times governments address this
07:31 by collecting information from third parties
07:34 that they can now use to cross check what the individual is declaring to them.
07:37 And third is that beyond knowing what is owed,
07:40 government needs to be able to collect these revenues into the state coffers.
07:43 There needs to be ways of,
07:44 on the one hand,
07:45 facilitating people to pay,
07:47 so making it easy and lowering compliance costs,
07:49 but then on the other hand,
07:51 to,
07:51 um,
07:52 bring the force of the law
07:53 on people who do not wish to comply.
07:55 And so these three identification,
07:57 detection,
07:57 and collection are things that we'll come back to.
08:01 For the paper today we're going to be highlighting lessons from
08:04 evidence
08:05 in the literature over
08:07 the last decade or
08:09 so on
08:11 interventions to increase tax capacity and what we've
08:13 learned about governments being able to collect more revenues
08:16 and the first pillar that I would examine is technology.
08:19 Technology,
08:20 as in every other face of life,
08:22 has delivered very powerful tools which I'll show you have created much,
08:26 a lot of potential for identification,
08:28 detection and collection.
08:30 Next we move to the role of tax officials because technology is a tool.
08:34 It has to be used by someone,
08:35 and tax officials
08:37 are the key elements of tax administrations,
08:40 and we'll be looking at specifically the way in which they're deployed
08:43 and the ways in which they incentivize and how this affects their performance.
08:48 What I'm going to argue is that
08:50 we have enough evidence of very effective
08:54 interventions on
08:56 the use of technology and the deployment of tax officials
08:59 which leads to the question of why don't we see more of this
09:02 and for that we'll turn to the third point which is
09:04 politics
09:06 and the argument is that
09:07 political incentives at the end of the day are what
09:10 determine the level of investments that states are going to pursue
09:13 in these different elements of tax capacity.
09:16 The level to which a government believes that it
09:18 needs taxes and the costs and benefits of pursuing taxation
09:22 are going to eventually shape
09:23 the
09:24 tax
09:24 investments.
09:26 So moving to the first
09:28 point which is on technology,
09:29 uh,
09:30 there are now quite a wide variety of tools
09:33 that governments can use in increasing tax capacity.
09:36 So starting from identification,
09:37 now it's possible to have things like biometric IDs to identify citizens
09:42 and even things like property databases that are all geocoded for easy tracking
09:47 on the detection side,
09:48 there's a.
09:49 The new tools which are very exciting,
09:51 um,
09:52 for instance,
09:52 things like electronic fiscal devices which
09:55 are installed in
09:57 stores and retail stores,
09:58 and as soon as they're checking out customers,
10:00 that data is immediately sent to the tax authority
10:03 so that they know exactly what the revenues are.
10:05 We also have things like electronic payments through credit card or mobile
10:09 money that provide again a paper trail of
10:11 the transactions that are occurring with the firm.
10:14 We have things like electronic invoicing that are used for the VAT,
10:18 so it's very useful in cross checking
10:21 the um reports of a transaction between a seller and a buyer,
10:24 and Ann is going to um describe this more.
10:27 We also have um.
10:29 Electronic um systems for data transfer across countries that can be used to track
10:34 offshore wealth and,
10:35 um,
10:36 Pier is going to discuss more about um
10:38 monitoring evasion across countries
10:41 on the collection side,
10:42 um,
10:42 there are many tools now like electronic
10:44 tax transactions for instance like electronic filing
10:47 and payment that all serve this facilitation tool to make paying taxes easier.
10:52 I am,
10:52 I went through this very quickly because I covered some
10:54 of this in the previous PRT that I did in 2021
10:57 and I also have um references to papers that go more in depth.
11:01 So while we do have a
11:03 number of studies now that show in very nice um rigorously identified um settings
11:08 um
11:10 the impact of technology on revenues,
11:13 it's nice to also see this across countries so here we have
11:16 um.
11:18 Relationship between the use of technology in the country.
11:21 So this is a digitalization survey conducted by
11:23 the OECD in 2023 where they ask countries
11:27 all the different types of technologies that they have,
11:29 and we can see that this positively correlates with tax revenues as a share of GDP
11:33 even when you control for the income level of countries.
11:36 So this is just nice data to see on the aggregate.
11:39 And now once
11:40 I would just like to emphasize on the use of technology is that
11:44 it's not enough for countries to adopt
11:46 a nice technological system and we have this new databases.
11:49 There needs to be complementary actions
11:51 that are taken to actually maximize the use of this
11:54 of it.
11:56 So
11:56 to take one example,
11:58 the benefits for many of these electronic systems is that a
12:02 lot of high quality tax data is generated in the process.
12:06 What we find is that this is often.
12:07 Underutilized
12:08 in
12:09 many low income countries
12:11 perhaps due to lack of skills that's a major thing,
12:13 but there might be other constraints
12:14 and so being able to systematically use this data
12:17 for detection and for enforcement
12:20 is really important
12:21 and
12:22 in the absence of this we've seen studies where a lot of money is left on the table
12:26 um
12:26 and so one example of the ways it can be used is to be able to use it to inform,
12:30 even,
12:31 um,
12:31 to inform a risk-based audit system to direct enforcement efforts.
12:37 A second
12:38 important practice is that
12:41 Governments need to realize that taxpayers are smart,
12:43 you know,
12:44 time after time again you find that once the government comes in one direction
12:48 to,
12:48 for instance,
12:49 get more information on the revenues,
12:51 taxpayers are very strategic,
12:52 and then they
12:53 adjust on another margin.
12:54 So if the government knows that I'm
12:55 collecting more revenues,
12:56 then I remember that I have all these expenses
12:59 that I also need to add onto my,
13:00 my tax declaration.
13:01 And so
13:02 being able to close loopholes and have a comprehensive view of the,
13:05 um,
13:06 financial situation of firms is important.
13:09 And third,
13:10 which I would probably emphasize the most,
13:12 is that technology is not going to all of a sudden
13:15 solve
13:15 all the tax enforcement problems.
13:17 Instead,
13:18 it's a tool that must be
13:20 coupled
13:21 with the traditional enforcement.
13:22 I will refer to a project that I did with the Liberia Revenue Authority
13:27 in Monrovia where
13:29 they invested in creating a new property database and so
13:32 collecting all this information on where properties were located,
13:34 who the owners were.
13:35 And then we had a randomized experiment where we sent out these
13:39 notices to people
13:40 with its detection information to let them know that the government
13:43 knows who you are and knows that you've not paid your taxes,
13:46 and what we find is that
13:47 that information on its own does not actually move compliance behavior.
13:51 The only thing that moves it is when it's coupled
13:54 with information on penalties and enforcement as well.
13:56 And so technology on its own,
13:58 it has to be in a framework of
14:00 the traditional legal enforcement tools.
14:03 Now I'm moving to the next um
14:05 component which is the tax officials.
14:09 So tax officials,
14:10 what we see is that they are a core input to tax capacity.
14:14 So in this graph I have on the x axis the
14:17 population
14:18 per staff.
14:19 This is from the International Survey and Revenue Administration,
14:22 and what we find is that there are a lot more staff.
14:25 Per population in
14:27 higher income countries and higher tax collecting countries,
14:31 so for instance,
14:32 in the United Kingdom you have one tax official for about every 1000 citizens,
14:37 and this is an order of magnitude higher,
14:39 like 1 per 10,000 in a place like Ghana.
14:42 And not only do they have more staff,
14:44 but they also use them differently,
14:45 so they tend to more of
14:48 people in higher income countries are assigned to things
14:50 like audits and investigations compared to lower income countries.
14:54 Beyond the numbers and incentives,
14:55 we also see things along the deployment and incentives of
14:59 these tax officials.
15:00 So we find,
15:00 for instance in the case of Indonesia that
15:02 being able to
15:04 assign
15:05 more staff to higher tax bases
15:08 from things like medium tax offices and large tax offices
15:11 that actually increased revenue collection by about 50%.
15:15 We also have studies showing that prioritizing enforcement of
15:18 the areas with the highest likelihood of yielding revenues and
15:22 showing um another paper on the assignment of teams,
15:25 so pairing
15:26 high
15:26 capacity team members together,
15:29 that also
15:30 in a cost neutral way increases the amount of taxes that are collected.
15:35 And now moving on to the incentives,
15:36 we have evidence both of these case papers actually from Pakistan showing that
15:40 um either financial incentives or even non-financial incentives
15:44 like having a choice of where you're posted to
15:46 lead to higher performance by tax officials.
15:50 And so in this too um
15:51 thinking about technology and human tax agents,
15:54 a key
15:55 task is really balancing the strength of both
15:58 when in in lower income countries as they
16:00 start to adopt more tax more tech based systems
16:03 so for example
16:04 we.
16:05 Know that technology-based systems,
16:06 it creates the standardized interface for everyone.
16:09 It processes mass massive amounts of tax data,
16:12 but then the downside is that it might also disadvantage certain
16:16 taxpayers,
16:16 those that are lower income,
16:17 lower education that have more difficulty accessing the system.
16:20 On the side of the tax officials
16:23 we find that
16:25 from this in a lot of in-person interactions that they
16:27 have they actually gain a lot of rich contextual knowledge
16:30 of the tax base
16:32 that could be lost
16:33 if in the transition to online systems and so being
16:36 able to balance this advantage that they have with the risk
16:39 of collusion and bias and extortion that could
16:42 come from these interactions is an important um
16:45 uh
16:46 process that countries need to go through.
16:49 In thinking about the um.
16:52 The
16:53 interventions that I've described so far,
16:54 one thing we do in the paper is we study,
16:57 um,
16:58 papers that have been published since 2011
17:01 and look at
17:02 a range of interventions that have
17:05 increased tax collection substantially.
17:07 So we have 39 interventions,
17:08 26 studies,
17:09 a wide range of countries,
17:11 and this is what we have.
17:13 So
17:14 these are the different categories that I've just described
17:16 with the incentives of tax officials,
17:18 their deployment,
17:18 identification,
17:19 capacity,
17:20 detection.
17:21 Collection capacity specifically in the facilitation
17:24 arm as well as the enforcement
17:27 and the two main takeaways from this table
17:30 and one is that
17:32 there's now
17:33 a lot of evidence on things that
17:35 work that increase tax collection quite substantially,
17:37 you know,
17:38 so for instance we look around um the middle row of third party detection
17:42 we see several studies that are yielding
17:45 um.
17:46 You know,
17:46 30,
17:46 40,
17:47 50% increases in
17:49 tax revenues
17:50 and the darker shaded colors are showing that this is happening on a very high base.
17:55 So the bases here are over $100,000 US
17:58 and so this translates to pretty substantial increases.
18:03 Second thing to note is that some of the areas
18:05 where we're seeing the biggest increases also seem to be around
18:09 the officials of the tax
18:11 incentives,
18:12 the incentives of tax officials,
18:13 and the enforcement.
18:15 And so
18:16 this is really
18:17 going back to the point I was making earlier that we
18:20 have seen a lot of interventions that can yield increases.
18:23 And so why don't we see more of this?
18:25 And this leads to the role of politics.
18:28 And the fact is that
18:30 politicians might rationally be unwilling to tax
18:34 because taxation
18:36 uh
18:36 can very easily have a political cost.
18:39 And so for instance we've seen evidence that asking citizens
18:42 to pay taxes increases the level of political engagement.
18:45 In fact,
18:45 we see cross country evidence that
18:47 when competitive elections are approaching governments
18:50 actually reduce their tax collection effort.
18:53 And so one thing to really understand is
18:55 what are the factors that shape the willingness of governments to pursue taxation,
18:59 and we in the paper go over three main
19:02 things which is the availability of alternative revenue sources,
19:05 um,
19:06 the level of political competition,
19:07 as well as the available technologies that can reduce the salience of collection
19:12 of,
19:13 um,
19:13 of taxes.
19:15 And um
19:16 this is a case study
19:19 of Lagos,
19:19 Nigeria that really illustrates these different points I just mentioned.
19:24 So,
19:25 um,
19:26 this was in the early 2000s.
19:29 There was in Lagos,
19:30 the governor was from the opposition party
19:33 and there was a lot of political
19:34 competition between Lagos and the central government.
19:36 And what we have is that at that time
19:39 there was a dispute about local government and so the federal government cut off
19:43 the monthly
19:45 oil revenue allocation
19:47 to these local governments in Lagos state.
19:48 So the tap that was flowing all of a sudden dries up.
19:51 And what we find is that
19:53 even though we can already see an upward trend before that happens,
19:57 this ramps up significantly
19:59 when there is no longer the oil money flowing.
20:01 And so
20:03 the question is
20:04 how did the government achieve this.
20:06 There were,
20:07 there's a lot of papers in the political science literature that actually
20:10 studied this in depth and some of the things that point to are
20:14 investment in technology
20:15 as well as investment.
20:16 In personnel
20:17 and how a lot of these um
20:19 were things that the government did to ramp up this tax collection.
20:22 One,
20:22 I was living in Lagos actually right in 2006
20:25 and so it was very it was so salient.
20:27 Everything was pay your tax,
20:28 pay your tax.
20:29 At the same time there was a lot of public good provision that was happening that,
20:32 you know,
20:33 roads were clean and it's like,
20:34 yay,
20:34 this is what your tax money is doing
20:36 and so it really was to me a very visual,
20:38 um,
20:39 experience and seeing
20:40 how this works in practice.
20:43 Now,
20:43 going forward,
20:44 there's,
20:45 um,
20:45 I think there's a very large research agenda ahead,
20:48 but a few things to highlight are things around
20:50 understanding these complementarities between
20:52 tax
20:54 technology and tax officials and how we can harness even the new tools like in
20:58 AI and um
21:00 other changes like digital currencies and being
21:02 able to really stay at the forefront
21:04 of this.
21:05 And second is really thinking carefully
21:07 about how to build broad-based political support
21:09 for strengthening taxation.
21:11 And on the World Bank side specifically,
21:13 we are very excited that we have a new initiative
21:17 to work in harnessing
21:19 electronic tax data that countries are producing using all the
21:22 new technology investments that they've been making over time,
21:26 and this is such a rich source of,
21:28 um,
21:28 and
21:30 and is going to go more into this in a very
21:33 good example of what this data enables us to do,
21:35 but in thinking about policy questions that can be answered.
21:39 And as I mentioned,
21:40 countries that
21:41 own this data many times underutilize them.
21:43 That's one of the things we're very excited about is also to be able to partner
21:46 with countries and providing technical support
21:48 and trainings and providing code and documentation
21:51 for how they can utilize such data on their own.
21:54 So
21:54 thank you all again for your attention and for being here.
21:57 I'll hand it over to Anne who's online.
22:03 Thanks,
22:03 Ayebola.
22:04 Um,
22:05 Anne,
22:06 are you?
22:08 Are you there?
22:12 We saw,
22:13 we saw you earlier,
22:14 but oh there you are.
22:15 OK,
22:15 great.
22:27 Great,
22:27 um,
22:28 thank you so much everybody for joining.
22:30 Thank you,
22:30 Dion and thank you Manuela for discussing our work.
22:33 Um,
22:33 I'm really excited to present to you our paper
22:35 on the value added of value added taxes.
22:39 So this is joint work with a great team of co-authors,
22:42 uh,
22:42 Julia Vedan Mazar,
22:44 and Miguel,
22:46 um,
22:46 and it was really a collaboration across several institutions,
22:49 um,
22:49 so.
22:49 So between the World Bank,
22:51 um,
22:52 the Tax Dev Center,
22:53 which is the Center for Tax Analysis in Developing Countries
22:56 co-delivered by ODI and the Institute for Fiscal Studies,
23:00 and ICTD,
23:01 the International Center for Tax and Development,
23:05 and most importantly,
23:06 we had a superb coalition with revenue authorities
23:09 in the 11 countries that we studied,
23:10 so we're super grateful for that.
23:13 Uh,
23:13 for those connected online,
23:14 um,
23:15 the dance,
23:15 my co-author is,
23:16 um,
23:17 also happy to answer any questions that you may have in the chat.
23:20 He has held,
23:21 uh,
23:21 delved more deeply than any of us in the team into
23:24 the VAT data and the VAT systems with our partner countries.
23:27 Um,
23:28 he can,
23:28 for instance,
23:29 say terms like VAT withholding in 5 different languages.
23:32 Um,
23:33 so he's great,
23:33 and you can ask him all the hard questions.
23:36 Um,
23:37 so
23:38 the VAT is arguably the single most important tax in,
23:43 uh,
23:43 lower income countries,
23:46 um,
23:47 and it's also the fiscal tool that governments often
23:50 turn to first in a time of crisis.
23:54 Uh,
23:56 it raises a lot,
23:57 uh,
23:58 about a third of total tax revenue and also has potential to,
24:02 uh,
24:02 further increase,
24:04 um,
24:04 the VAT,
24:05 or rather the textbook model of the VAT has two important features,
24:10 um,
24:11 that can make it an attractive policy tool.
24:14 The first of these features is what we call production efficiency,
24:18 so the fact that the VAT does not
24:20 distort input prices and firms' production decisions.
24:25 Uh,
24:25 the second feature is that,
24:27 uh,
24:27 the VAT,
24:29 is revenue efficient.
24:33 Because
24:34 it has this inbuilt structure that makes it hard to evade
24:37 and hence allows it to collect a lot of revenue.
24:40 This is because firms report both
24:43 their VAT liable sales and also their VAT liable purchases
24:47 to the government.
24:48 And so for each transaction,
24:50 the government gets two reports,
24:51 one from the seller and one from the buyer,
24:53 and these two reports can be cross-checked.
24:55 So the VAT lends itself very well to the
24:58 digital enforcement technologies that Oyebola has just mentioned.
25:02 Now,
25:02 despite these attractive features,
25:04 the performance of the VAT is often thought to fall short of the policy's potential,
25:10 and especially in lower income countries.
25:12 So,
25:13 let's look at the sea efficiency,
25:15 which is an imperfect but commonly used metric for VAT efficiency.
25:20 It's basically VAT revenue over the tax rate times consumption.
25:24 So asking how much could you potentially
25:26 be collecting if all consumption was taxed.
25:29 And the sea efficiency is generally quite far away from 100%,
25:33 um,
25:33 and is lower in lower-income countries.
25:37 Now,
25:37 this is an aggregate metric.
25:39 It doesn't necessarily tell us
25:41 how VAT systems work in detail and how they can be improved.
25:47 So,
25:47 to shed more light at
25:49 how VAT systems function,
25:51 we assembled microdata,
25:53 meaning uh firm level VAT data
25:55 from 11 countries at different income levels,
25:59 ranging from $500 US dollars per capita in Ethiopia to 450,000.
26:04 T00 dollars US dollar per capita in France
26:07 and we then use these data to establish four novel stylized facts on VAT systems
26:12 and using this type of data as part of a broader agenda
26:15 using micro tax data from a diverse set of countries to study,
26:20 uh,
26:20 important development questions.
26:22 Um,
26:22 so we have another paper with Pierre for instance and
26:24 co-authors where we study in a group of 15 countries,
26:27 uh,
26:27 effective corporate tax rates and the global minimum tax.
26:30 And we have another related paper with Pablo Garriga and co-authors where we
26:34 use the tax data to study the effect of COVID on firms.
26:39 So,
26:39 in this paper,
26:40 we document,
26:41 uh,
26:41 through the four facts,
26:44 how real-world VAT systems diverge from the textbook ideal of the,
26:50 uh,
26:50 the ideal textbook model of the tax.
26:53 Um,
26:55 now,
26:55 many governments or some governments have in the recent coup in the
26:59 last couple of years have toyed with the idea of replacing.
27:02 The VAT,
27:04 uh,
27:04 so in the end of the paper we then consider alternative tax instruments
27:07 and show that they actually could not do better than the real world VAT systems,
27:12 imperfect as they are.
27:13 So getting rid of the VAT is not the answer.
27:16 Instead,
27:16 improving the VAT is the answer,
27:18 and the World Bank and our partners at TaxDev and ICTD,
27:22 we're here to support countries in their endeavor to improving their VAT systems.
27:26 So now,
27:27 let's delve into the 4 facts.
27:29 Starting with fact number 1,
27:30 which is about the concentration of VAT revenue.
27:34 So,
27:34 in the textbook theory of the VAT one attractive
27:37 feature of the tax is that it is broad-based,
27:40 remitted in small chunks along the entire value chain,
27:43 so by many firms throughout the economy.
27:46 In practice,
27:46 however,
27:47 we find that VAT collections are extremely concentrated,
27:51 um,
27:51 with the largest 10% remitting,
27:54 uh,
27:54 about 90% of VAT revenue,
27:56 and the share is even larger in the lower income countries.
28:01 As a result,
28:02 tax authorities have to spend quite a lot of time dealing
28:05 with small businesses that ultimately account for a little revenue,
28:09 and the small businesses have to spend hours trying to comply with the VAT.
28:15 Now,
28:15 the tax practitioners wouldn't be surprised by the fact that
28:18 tax collections are generally so skewed towards bigger firms,
28:22 but the level of concentration and this negative correlation
28:26 with development are facts that we hadn't previously documented,
28:29 and it's really the newly available microdata in a diverse set of countries
28:33 that allows us to establish this point.
28:37 Fact number 2,
28:38 relates to effective tax rates.
28:40 So the ideal textbook VAT
28:43 avoids distortions to production by imposing the same
28:46 VAT rate on all goods and services.
28:49 In practice,
28:49 however,
28:50 VAT systems diverge from this ideal.
28:52 So exemptions and reduced rates are common with
28:55 often problematic implications for equity and efficiency.
28:59 So here using microdata,
29:01 we can show that the selling goods that are reduced rated or exempt from VAT
29:07 is actually more common among larger firms,
29:09 and the sales of these goods are also larger for larger firms.
29:13 Um,
29:13 and because of that,
29:15 effective VAT rates end up being lower for the sales for of larger firms
29:20 uh compared to the sales of smaller firms,
29:22 and larger firms are typically also those who
29:25 are catering more to the richer customers.
29:28 Now this result holds true even
29:30 within small,
29:31 uh,
29:32 industry bins.
29:35 Um,
29:36 now,
29:37 this is all focusing on the formal sector once you
29:39 consider informality that adds an additional layer of complexity,
29:42 and Pierre is going to discuss that in more detail.
29:46 Moving on to fact three,
29:49 we're looking here at non-reclaiming of input VAT.
29:53 So the fact that firms remit VAT on their output,
29:56 but can deduct VAT on,
29:59 uh,
29:59 the input,
30:00 the VAT that they paid on their inputs,
30:02 that's also a defining feature of VAT system and it's
30:05 also necessary to ensure that the VAT is production efficient.
30:09 However,
30:10 it is actually common for small firms to not claim any input.
30:14 VAT.
30:15 So in the countries we study,
30:17 uh,
30:18 almost 50% of the smallest firms,
30:20 so the 5,
30:22 the those in the 5% in the bottom 5% of the sales distribution,
30:27 those firms do not claim any input VAT.
30:29 And overall the share of firms not claiming
30:32 input VAT is higher in lower income countries.
30:37 Again,
30:37 this holds within,
30:38 uh,
30:39 industry bins.
30:41 Um,
30:42 and when firms do not claim any input VAT,
30:44 uh,
30:44 despite using taxable inputs,
30:46 then the VAT chains are broken,
30:48 and this introduces distortions and limits also the revenue efficiency of the tax.
30:54 Our last fact relates to VAT refunds.
30:58 Uh,
30:58 so refunds help underpin the destination principle of the VAT,
31:02 so the fact that the tax is ultimately levied where the good or service is consumed.
31:07 And so this requires that exports are zero rated,
31:10 which means that exporters
31:12 often generate refund requests.
31:15 Now refunds are a critical part of VAT systems.
31:18 Uh,
31:19 in higher income countries,
31:20 for instance,
31:20 about 30 to 50% of VAT revenue ends up being refunded back to taxpayers.
31:26 But in lower income countries,
31:28 VAT refunds are constrained by
31:31 government's limited uh administrative capacity,
31:34 the risk of fraud,
31:35 so firms submitting kind of fraudulent refund claims,
31:39 um,
31:39 and sometimes also governments reluctance to actually forego revenues.
31:43 And this means that firms often wait for months and months,
31:46 if not years,
31:48 to receive VAT refunds.
31:51 The time to receive a refund is again
31:53 strongly negatively correlated with the country's development level.
31:57 Now,
31:57 long refund delays have serious implications for businesses' cash flows,
32:02 and although most countries try to facilitate refunds for exporters,
32:06 um,
32:07 because
32:08 exporters are most likely to generate large refunds,
32:11 these facilitation strategies seem not to be seem to be insufficient,
32:16 um,
32:16 because in our countries,
32:18 in the data that we study,
32:20 exporters are actually not less likely to generate refunds than other firms.
32:25 Now,
32:26 given all these imperfections of real world VAT systems,
32:29 should countries
32:30 consider replacing the VAT by alternative tax instruments,
32:35 um,
32:37 such as,
32:37 for instance,
32:37 a retail sales tax or a turnover tax
32:40 that may be simpler to collect?
32:42 Now,
32:42 this sounds a bit
32:44 potentially crazy,
32:44 but it's actually been policy questions that have been debated,
32:47 uh,
32:48 and discussed in,
32:49 uh,
32:49 recent years in,
32:50 for instance,
32:51 Ghana,
32:51 Malaysia,
32:52 and Zambia.
32:54 So what we do is to simulate how these alternative tax instruments would work,
32:59 um,
33:00 focusing on the case of Pakistan,
33:02 which is one of the countries where we have
33:04 very detailed transaction level data that allows us to simulate these other taxes.
33:12 So in Pakistan,
33:13 actual VAT revenue
33:15 before COVID was about
33:18 ₹1500
33:20 billion.
33:22 Now,
33:22 um,
33:23 the data allows us to tag business to consumer sales,
33:26 so we can then ask
33:28 how much revenue would be collected by a retail sales tax,
33:32 so a tax that is only levied on the sales of retail firms.
33:38 Now in theory,
33:39 a retail sales tax is the same as a VAT.
33:42 The difference is just that the VAT is levied in small chunks along the value chain,
33:46 but the retail sales tax is entirely remitted by
33:49 the retail firms without any input credit mechanism.
33:53 This sounds kind of simpler and there's fewer firms
33:56 to to deal with for the tax administration,
33:58 but in most lower income countries,
34:00 as in Pakistan also retailers are much smaller
34:03 firms and less tax compliant than upstream firms,
34:06 say manufacturers,
34:08 um,
34:08 and so a retail sales tax,
34:10 even if levied at the same
34:12 rate as the VAT,
34:14 so 17%.
34:15 would collect less than 1/3 of the revenue that the VAT collects,
34:21 so that doesn't really seem to be a viable option.
34:24 Now,
34:24 alternatively,
34:26 the government may say,
34:27 OK,
34:28 let's collect the tax from all firms along the value chain,
34:31 but let's just get rid of the input tax credits.
34:33 That's complicated.
34:34 Let's just make it a turnover tax.
34:36 So it's just a
34:37 tax on the total sales of all firms,
34:40 no input tax credit.
34:42 Sounds also simpler,
34:43 right?
34:44 Um,
34:45 so,
34:45 here,
34:45 what we can do in our data is to calculate what would
34:48 be the rate at which the turnover tax has to be applied
34:52 to collect the same revenue as the VAT holding everything else constant.
34:58 Now,
34:58 that counterfactual turnover tax rate is 6.1%.
35:02 So that's a very high.
35:04 Very distortionary
35:06 uh
35:06 rates because of this turnover tax itself is distortionary as it cascades through
35:11 the value chain.
35:13 Now,
35:14 as the turnover tax is distortionary,
35:16 firms' output with such a high turnover tax rate might actually go down.
35:20 So it's quite possible that the government couldn't even collect as much
35:24 revenue as it can with the VAT using a turnover tax.
35:29 So in conclusion,
35:31 real world VAT systems diverge substantially from the textbook model,
35:36 uh,
35:37 but the VAT is ultimately irreplaceable.
35:40 So how can countries improve their VATs?
35:43 Well,
35:44 they can reduce exemptions and reduce rates,
35:46 especially if those have little equity benefit
35:48 or even make the tax system regressive,
35:51 um,
35:51 and also facilitate refunds using the
35:54 technologies that Oyebola has just discussed.
35:57 Um,
35:58 governments can also task larger firms
36:01 or digital intermediaries with withholding the VAT
36:03 from smaller firms and thereby improving their
36:05 compliance and potentially lowering compliance costs.
36:09 Um,
36:10 but ultimately we need more analysis to better
36:12 understand the puzzling compliance behavior among small firms.
36:17 That may be simultaneously kind of evading taxes but also leaving
36:20 on the money on the table by not claiming input tax credits
36:23 and we need more work uh on optimizing the use of tax data
36:27 for enforcement and facilitation,
36:29 taking into account also spillover effects in networks
36:33 and in general I think our
36:35 paper suggests that there's a lot we can learn from
36:37 microdata but also a lot we still need to learn.
36:40 Um,
36:41 to better understand tax systems and their equity and efficiency implications,
36:45 um,
36:46 and these microdata,
36:47 which are now becoming more and more available,
36:49 allow us to shed light also at
36:50 other important dimensions of the development process,
36:52 such as,
36:53 for instance,
36:53 firm growth and labor market dynamics.
36:56 And so this is why we're all really excited to
36:58 be soon launching Datax as a lab to study taxes,
37:02 firms,
37:02 and jobs using micro-tax data.
37:05 And uh Datax would be a collaboration between EE,
37:09 DC,
37:09 and other parts of the bank that work uh in this space
37:12 and together also with external partners such as
37:15 the Tax Dev Center and the EU Tax Observatory
37:18 and of course most importantly working with ministries
37:20 of finance and tax administrations around the world.
37:24 So stay tuned and thank you so much for joining today.
37:30 Thank you,
37:30 Anne.
37:31 I'm amazed that
37:32 the day tax folks managed to get the towel in there.
37:34 That that that's so,
37:35 that's so nerdy.
37:38 OK,
37:39 over to you,
37:39 Pierre.
37:43 Thank you Dion.
37:44 So thanks Anna Mayevola.
37:46 This was really interesting.
37:47 Uh,
37:48 Ryan,
37:48 the slides are not showing yet.
37:52 So this last talking is on the dimension of tax equity,
37:55 right?
37:55 We saw administration we saw efficiency of the
37:58 key instrument and now we're gonna talk about the
38:00 distribution of part of taxation
38:03 so
38:04 first win a context as we know a lot of governments are feeling a lot
38:07 of fiscal pressure but at the same time
38:09 there's a very high income concentration that has
38:11 actually if anything worsened in the last 20 years
38:14 and we know that this needs to be addressed multidimensionally.
38:17 On market income side we need active labor market policies,
38:21 you know,
38:22 skills building,
38:23 we need also targeted transfers and flexible social insurance program,
38:27 but we also need a third leg which
38:29 is progressive and fair taxation and by progressive tax
38:32 I'm gonna think of having higher tax rates on those who are able to pay more.
38:39 So this presentation is based on the 3rd article
38:42 in the Journal of Economic Perspective called Tax Equity in
38:44 Law in Middle Income Countries which is co-authored with Lucy
38:46 Gadden and Anders Jensen plus some other ongoing work.
38:52 So the first thing we do in this paper is just set a framework of what is tax incidents,
38:55 who is actually paying taxes?
38:57 There's three key ingredients that someone needs to know
39:00 to think of tax incidents.
39:02 The first one is the easy one
39:04 what's the statutory incidents?
39:05 If we have the tax laws,
39:07 we can go and check what are the rates,
39:09 what are the base rules,
39:10 and so on.
39:12 The second one that's harder is the de facto incidents in practice
39:16 in a context where you have a lot of tax evasion,
39:18 informality avoidance,
39:20 who really pays the taxes
39:21 could be very different from who the law intended should be paying the tax.
39:26 For a long time this had been overlooked
39:28 in part because there was this optimistic assumption that maybe
39:31 this was not first order for high income countries.
39:34 In the last 10-20 years,
39:35 a lot of work has tried to correct this
39:37 and just to give you a sense,
39:39 you know,
39:39 these are graphs you probably know,
39:41 but the size of the informal sector.
39:44 We can take a proxy self-employment as a proxy of labor informality and
39:48 you can see here how how large it is and how it relates
39:51 with the country GDP per capita
39:53 and on the right hand side uh panel you have a more novel measure which is
39:56 consumption informality where we proxy this with the
39:59 share of consumption occurring in the traditional sector,
40:02 small stores,
40:03 self productions,
40:04 and the like.
40:07 So once we take,
40:08 you know,
40:08 these facts into account,
40:09 we know that de facto incidence is gonna matter for equity too.
40:13 The third piece is arguably the hardest,
40:15 and that's somewhere where we need a lot more work,
40:17 which is economic incidents.
40:18 Market prices are gonna respond to any tax,
40:21 potentially shifting the burden away
40:24 from those who are remitting,
40:25 paying the tax
40:26 with those who are trading with them.
40:28 An example,
40:29 for example,
40:30 is comes from the sales tax.
40:32 A sales tax is remitted by a firm but very often
40:35 is passed on to consumers in the form of higher prices.
40:39 And so you know,
40:39 while that had been an active literature in OECD countries,
40:42 it's more recent that we see some paper in low and middle income countries,
40:45 for example,
40:45 the work of our colleague,
40:47 uh,
40:47 Dario Tortarolo,
40:48 where is he?
40:49 I saw him earlier,
40:52 um.
40:54 So
40:56 The second chapter then think of given the current
40:59 tax revenue and tax mix of developing countries,
41:02 what is the likely
41:04 incidence of these taxes,
41:05 in particular those first two factors,
41:07 right?
41:07 The statutory and the de facto incidents.
41:10 I call this a tale of two taxes because really with two
41:13 main instruments you have most of the tax revenue in developing countries.
41:16 On the one side we have indirect taxes,
41:18 of which by far the main component is the VAT.
41:22 And on the other hand we have the personal income taxes.
41:25 In developing countries,
41:27 indirect taxes are about 60% of tax
41:29 revenue have excluded social security from here.
41:33 Well,
41:33 the personal income tax is around 10 to 15% of tax revenue,
41:37 and the real big shift that we observe across development.
41:41 Is that the personal income tax starts taking off but at relatively
41:44 high income levels right after 15-20,000 uh in GDP per capita.
41:49 For rich countries,
41:50 eventually the personal income tax represents about 50%
41:52 of tax revenue and really displaces indirect taxes.
41:59 So what is the progressivity of the PIT in developing countries?
42:02 So first,
42:02 let's start with what we've just seen,
42:04 which is that developing countries are characterized with a
42:07 large share of activity in the informal sector.
42:10 A high reliance on indirect taxation.
42:13 And low levels of personal income tax collection,
42:16 so this does not bode very well when we think of equity right at baseline.
42:20 Now maybe
42:22 the low PIT is still very progressive if it's all born by the very top
42:26 maybe that's fine we still have progressivity in the tax system.
42:30 So in work we've done with colleagues in
42:32 the poverty group Matt Waipoi and Mariano Sosa,
42:34 we try to get the de facto distribution incidence of
42:37 direct taxes based on individual about 100 CEQ country studies.
42:42 And so here
42:43 when we compare high income countries to developing
42:45 countries you see again that level point,
42:47 right?
42:47 The high income country at all these sides of the income distribution lies above.
42:51 But more importantly,
42:52 you see the gap in progressivity,
42:55 right?
42:55 So let's compare for example the top side of
42:57 high income countries relative to the bottom side.
43:00 This is a gap in average tax rate of around 15% points in high income countries,
43:05 and you can see it's only 5 or 6% points
43:08 in developing countries,
43:09 right?
43:10 So much less progressivity is brought from this
43:13 relatively weak personal income tax.
43:19 So optimal taxation,
43:20 you know,
43:20 in this scale is in this context with constrained income taxes
43:23 is kind of a newer problem for the literature because if
43:26 we look at the textbook policy it would say all redistribution
43:29 should be achieved with the income tax and some transfers.
43:32 In practice we see that when you have such a weak income tax
43:35 you might still care about equity and think of other instruments to alleviate,
43:39 you know,
43:39 the lack of tax progressivity.
43:42 In practice,
43:43 what we see is that many countries actually do rely for
43:45 example on the VAT and indirect taxes for equity purposes too,
43:49 a very,
43:50 you know,
43:50 ubiquitous policies to have
43:52 zero rate or reduced rate on necessity goods think of food,
43:56 think of medicine and other products
43:58 sometimes also higher rates on luxury items.
44:04 So how should we think of introducing progressivity,
44:07 you know,
44:08 behind this kind of bleak picture I've just described,
44:11 and I think that,
44:11 you know,
44:11 there is cause for a lot of hope and progress in the next decade
44:15 here we might want to separate,
44:16 I think,
44:17 countries into two groups.
44:18 In middle income countries we can do a
44:19 lot more to strengthen the personal income tax,
44:22 and I'm gonna finish talking about this.
44:25 In lower income countries it's probably hard to do a lot with income tax due
44:29 due to some real information issues you
44:31 know Ebola had discussed the enforcement problem.
44:34 There's still two things,
44:35 two types of decisions that still matter a lot for equity.
44:37 The first one is just tax administration,
44:40 something we rarely think about for equity,
44:41 but it actually matters potentially a lot.
44:44 We dedicate a whole chapter in the
44:45 paper with examples coming from property taxation.
44:47 Property taxation,
44:48 if done well,
44:49 should be quite a progressive tax.
44:51 We know that real estate is concentrated at the top typically.
44:54 But for example in a paper
44:55 in Senegal,
44:56 Knebelman,
44:56 Pulikan,
44:57 and Sar
44:58 find that when it was left to discretion
45:00 this ended up being kind of a
45:02 neutral tax in terms of progressivity
45:04 because of where the agents who choose to enforce how they value property.
45:09 Upon,
45:09 you know,
45:09 much better automation property tax evaluation that improved
45:13 a lot the progressivity of the property tax.
45:17 The third element
45:19 is that when your income tax is constrained,
45:21 you really want to think of the design of the
45:23 VAT and of indirect taxes and so I'm gonna show you
45:26 a few examples of how that matters.
45:29 So this is work from another paper with Lucy Gedon and Anders Jensen
45:34 uh
45:35 where we've taken household surveys in about 33 countries.
45:39 And
45:40 We've plotted the informal budget share as a function
45:44 of a household's income,
45:45 so let's take the example of Mexico.
45:47 What this is showing you is that for household in the bottom side,
45:51 almost 60%
45:53 of their consumption occurs in those traditional stores or from self-production.
45:57 When you go to the top Dile,
45:59 this falls to around 20%.
46:01 OK,
46:02 so very different distribution patterns in the way household shops,
46:06 right,
46:06 consume,
46:07 which can have important consequences for equity.
46:09 So what does that mean?
46:12 First,
46:12 consumption taxes,
46:13 once you take into account the informal sector,
46:16 are more progressive than often considered.
46:18 Now how much more progressive,
46:19 you know,
46:20 we estimate around a drop of 1% point of Gini,
46:24 a 2% drop.
46:26 It's not nothing,
46:27 it's not huge either,
46:28 right?
46:28 The personal income tax in rich countries,
46:30 it's about a 7 or 8.
46:32 Percentage points drop in genuser comparison.
46:36 The second one and maybe more relevant for policy
46:39 is that the policy of having low tax rate on necessity goods
46:43 is not
46:44 introducing a lot of progressivity once we consider the informal sector.
46:48 These exemptions benefit mainly people who consume informal stores
46:52 who tend to be the rich.
46:54 And so there's a key policy challenge which is how to phase those
46:57 exemptions out in a societally acceptable way and I think some of the work
47:02 uh of people of the bank is going this way.
47:05 The third one,
47:06 even more specific,
47:06 is that non-standard indirect tax policies could be desirable for equity,
47:10 for equity.
47:12 For example,
47:12 India has had for a long time Russian shops,
47:15 which is a subsidized price for fixed quantity
47:18 of basic goods.
47:19 And so Gaden shows that this raises welfare and access insurance in India.
47:23 In some recent work with David Being and Pula Ghosh,
47:26 we're looking at also higher rate on luxury items in the case of India,
47:29 we show that as expected those are quite progressive
47:32 but also not very distortionary
47:34 actually which is maybe more surprising.
47:37 And as I already mentioned,
47:38 we need more evidence on the economic incidence of indirect taxes
47:42 in Mexico for example when we look at a VAT hike.
47:45 We find that consumer prices
47:47 in the formal sector respond by about 75 or 80%,
47:51 so they raised by quite a bit,
47:52 but not that prices in the informal sector also increase,
47:56 right?
47:56 So thinking about kind of the supply chain effects,
47:58 competition,
47:59 and so on,
47:59 right?
47:59 And so we need more work
48:01 that thinks of the interaction,
48:02 for example,
48:02 between the formal
48:03 and the informal sector.
48:07 So the last part I wanna talk about,
48:08 you know,
48:08 a more
48:09 uh
48:10 positive view which is I think a lot can be done
48:13 in this coming decades
48:14 towards a progressive income tax in a lot
48:17 of countries and especially in middle income countries.
48:20 Because
48:20 improving,
48:21 you know,
48:22 tax equity will require more than just a better design VAT,
48:25 property taxes,
48:25 and targeted enforcement,
48:27 we will need progressive income taxes and some taxation of capital.
48:31 Income.
48:32 Now traditionally observing the income of the rich
48:35 and that of large firms of multinational has been challenging,
48:38 there's a few big issues.
48:39 The first one is offshore income and wealth.
48:43 The second one is that a lot of
48:45 the rich
48:46 and the very wealthy
48:48 hold income in corporations
48:50 and the lines between labor and capital income can become very blurred.
48:53 You can consume within the firm,
48:55 use assets of the firm to obtain loans,
48:58 decide on the timing of realization of income,
49:00 all things that the tax administration would have difficulty dealing with.
49:04 And finally,
49:05 for a long time,
49:06 corporate taxes were quite high in the 70s and the 80s
49:09 and would act as a backstop for the personal income tax.
49:12 Even if personal income tax was low,
49:14 you would collect it from the firms.
49:16 The
49:17 potential of the backstop of the corporate tax has
49:18 really been eroded via tax competition and aggressive profit shifting
49:22 in
49:23 the last two or three decades.
49:26 I wanna show you some work we've done with the EU Tax
49:28 Observatory part as part of the Atlas of the offshore world.
49:32 Those are a set of maps that are trying to
49:34 put numbers on tax eva avoidance and tax evasion.
49:37 Here I'm showing you,
49:38 for example,
49:38 a map of how much financial wealth is held abroad for each country.
49:43 The world average is around 12 to 13%.
49:46 This is a large number,
49:47 right,
49:47 of money held abroad.
49:49 This does not have to always correspond to tax evasion,
49:52 but even in a country like Denmark or Norway,
49:54 we estimate that only about 20 to 25% is reported domestically,
49:58 right?
49:58 So
49:59 that doesn't bode well,
50:00 I think,
50:00 for other countries.
50:02 You can see that Africa is particularly affected
50:04 with around 18% of its GDP held abroad,
50:07 but so is Latin America with around 14%.
50:12 Another map is how much
50:13 missing profits there are.
50:15 Those are profits,
50:15 you know,
50:16 profit shifting profits booked in tax havens and not
50:18 in the country where really the activity is happening.
50:21 The world average there is around 10 to 11% of corporate income tax missing globally
50:26 and again with countries in Latin America particularly affected there.
50:36 The policy response of the past two decades has been inadequate.
50:39 It's been to lower taxes on mobile basis,
50:41 kind of more tax competition,
50:43 if you will.
50:44 Lower taxes on capital income,
50:46 top earners,
50:46 and replace those with payroll and indirect taxes hurting equity in the process.
50:51 The second type of policy has to be to limit rate differential across tax types,
50:55 thus limiting the progressivity of the entire tax system.
50:58 So think of the flat tax rate in a lot of Eastern European or Central Asian countries,
51:02 right?
51:02 Low flat tax rates.
51:07 And so what that has meant that even in rich countries at the very top,
51:10 and here I really wanna highlight
51:12 that we are at the very,
51:13 very top,
51:13 we're talking about 100
51:15 millionaires or billionaires,
51:17 right?
51:17 The effective tax rate paid by individuals has fallen,
51:21 and this is when we use this comprehensive measure of income
51:24 that is also the income that is held within firms,
51:26 the undistributed profits of firms that are linked and attributed
51:30 to individuals.
51:31 And we can see that here we have,
51:33 you know,
51:33 a real program
51:34 of equity even in countries that have quite
51:36 high tax rates like France and the Netherlands,
51:38 the billionaires end up paying only about half as much taxes
51:41 as the rest of the population.
51:43 This is the type of exercise we can increasingly do by merging
51:47 with shareholder beneficial ownership data,
51:50 individuals and the,
51:51 the corporations they directly control.
51:53 So we,
51:53 for example,
51:54 with Thiago Scott,
51:55 um,
51:56 Tatiana Flores,
51:57 uh,
51:58 Santiago Esteros,
51:58 and Gabriello Kelly
52:00 of the World Bank,
52:01 we've tried to do the same type of exercise in Honduras,
52:03 for example.
52:04 And so here you can see
52:05 the first point to the left
52:07 is that at the very top all income is basically in undistributed corporate profits.
52:13 In the case of Honduras,
52:14 the drop is only at the very,
52:15 very top and maybe not as
52:18 spectacular as the one we saw,
52:19 uh,
52:20 in France and Honduras in part because the
52:21 corporate income tax stays relatively high in Honduras,
52:25 but we think this is really the type of work
52:26 we need to repeat in other countries to get a really
52:29 good distributional picture
52:31 of income taxation.
52:34 And so I want to finish with two messages.
52:36 The first one is a lot of people would think,
52:37 well,
52:37 taxing capital,
52:38 taxing the rich,
52:39 that's really bad for innovation,
52:41 but one thing we see repeatedly.
52:43 It's not just that the average tax is low,
52:45 is that there's very large dispersions in the tax rate
52:48 paid by the rich and paid by the very large multinationals
52:51 within an income group,
52:52 if you want,
52:53 within a size bin of firms.
52:54 Some firms are paying,
52:55 you know,
52:56 almost zero% tax rate while others are paying the full legal statutory tax rate.
53:00 This cannot be good for efficiency or competition,
53:02 right?
53:02 So that's something that we think is important to keep in mind.
53:05 The second one is this cause for some cautious optimism because of
53:09 what's been called the ABC of taxation of international taxation.
53:13 The A stands for automatic exchange of financial information.
53:17 The days of bank secrecy,
53:18 you know,
53:18 where you could put money in a Swiss bank account,
53:20 those are really over and that's good news.
53:22 This is quite recent and a lot of developing countries haven't made
53:26 most
53:27 of what they could do by exploiting such data for,
53:30 uh,
53:30 audits.
53:31 The second is that beneficial ownership registries
53:34 that link individuals to the firm they control
53:37 are being implemented or actually active now
53:39 in a majority of countries.
53:41 And the third is since this year January of 2024,
53:44 there's a corporate minimum tax
53:46 of 15%.
53:48 It's a global agreement we're gonna see how it actually plays out,
53:51 but
53:51 this would now put a floor on how much profit shifting could occur
53:55 globally.
53:56 Yet governments are gonna need assistance in this fast changing environment,
53:59 you know,
54:00 to track the evolution of this comprehensive top incomes,
54:03 to use the new data
54:04 to reduce evasion and avoidance at the top,
54:07 but also for advice
54:08 on the tax policy,
54:09 for example,
54:10 all the tax incentives that exist are now are gonna be impacted
54:13 by the corporate minimum tax.
54:14 And so this,
54:15 you know,
54:15 you've probably tired of here of day tax.
54:17 We hope to,
54:18 to do this as part of the day tax agenda
54:20 with Ano Yebola and many other colleagues in this room,
54:23 and we'll also be selling merchandise at the exit door.
54:26 Uh,
54:27 we have some hats,
54:28 no.
54:30 And so I'll leave it here.
54:31 Thanks a lot for attending this talk.
54:35 Thanks Pierre.
54:35 I mean that was a fascinating collection of,
54:37 of,
54:38 of,
54:38 of stylized facts,
54:39 policies,
54:40 programs,
54:41 and reflections.
54:42 Wow,
54:43 amazing
54:44 collection of three presentations.
54:46 Over to you,
54:46 Manuela for some uh reactions and reflections.
54:50 Thank you.
55:01 And uh what is uh I just moved it here.
55:04 Ah yes,
55:05 OK.
55:13 OK,
55:14 yeah,
55:14 yeah,
55:14 I can do that.
55:15 I just,
55:16 OK,
55:16 I just,
55:17 OK,
55:17 let's see,
55:19 yeah,
55:19 here it is,
55:20 OK,
55:21 um.
55:22 Good afternoon everyone and good evening for some that are connected.
55:26 Uh,
55:27 I want to thank you for inviting me to be here today.
55:30 Um,
55:31 uh,
55:31 I was very impressed with the work that was presented by Yebola and and Pierre.
55:37 Uh,
55:38 your work is very valuable to us in operations
55:41 and,
55:42 uh,
55:42 we hope,
55:43 I hope you will continue working together because I think,
55:46 uh,
55:46 you have very,
55:47 uh,
55:48 relevant work for the work we do directly with the clients.
55:51 So I will talk more about
55:53 how I see the work you are doing and how that
55:56 can help us in dealing with when we deal directly with clients
56:01 uh what are we doing in the regions and
56:03 the center because actually the department I lead is a
56:07 a center that then supports the regions
56:09 and how we can work together.
56:13 So,
56:14 um,
56:15 as you know,
56:16 we have been working on domestic resource mobilization for decades.
56:20 I joined the bank in 2005 and we were working on
56:25 domestic resource mobilization.
56:26 I did the PR in 2005 or 2006.
56:30 So why are we,
56:31 uh,
56:32 elevating this agenda?
56:33 Why are we talking so much about PRM?
56:36 So as you know,
56:37 as part of the evolution roadmap,
56:40 um,
56:40 DRM is one of the key components.
56:43 So the,
56:44 the,
56:44 the evolution road map has three parts basically.
56:48 One is that
56:49 we need to optimize the financial models of the World Bank
56:53 so then we can squeeze out more money for development.
56:56 Second is that we need to bring in the
56:58 private sector because public money is not enough.
57:02 And third is
57:03 domestic resource mobilization.
57:05 There is the realization
57:07 that the client countries,
57:09 that's why we call them in operations,
57:11 need to do their share.
57:13 There is not enough taxpayers' money and contributions from ODA.
57:18 There is not enough interest on the private sector
57:21 to finance development.
57:23 And just to give you some numbers,
57:26 so as you know,
57:27 uh,
57:27 and you have these numbers yourselves,
57:29 78% of low income countries and 56% of low mixed
57:36 collecting 15%
57:38 tax revenue,
57:39 tax revenue to GDP below 15%,
57:42 um,
57:43 as you know,
57:43 as per the literature,
57:45 its country to fully function and to do,
57:48 uh,
57:48 to provide service delivery in adequate manner should collect at least 15%.
57:55 More than 50% of the low income countries are
58:00 at high risk of debt stress or in debt distress already,
58:05 and they are
58:07 in this situation.
58:09 Because
58:10 they ran large fiscal deficits for a long,
58:13 long time,
58:14 we have,
58:15 we have done an analysis why countries are in
58:17 such a difficult position when it comes to debt,
58:20 and it was because
58:22 they didn't,
58:23 um,
58:23 consolidate on the spending side and they
58:26 didn't collect enough on the revenue side.
58:29 Just to give you another data point because this is very important,
58:33 I,
58:33 I think.
58:34 This year in 2004.
58:38 Country low income countries
58:40 will need to make repayment,
58:42 debt repayment in the amount of $175 billion.
58:48 This is mostly domestic debt,
58:49 Mark,
58:50 a domestic debt.
58:53 This is equivalent,
58:54 this amount is equivalent to 7.7.5% of the combined GDP of low income countries.
59:04 On average,
59:05 low income countries
59:07 spend less in education,
59:10 health,
59:10 and infrastructure than 7.5%,
59:13 so this is actually compromising
59:16 future growth.
59:19 Um,
59:20 We also did an analysis and we saw that
59:23 the countries right now face a situation of net.
59:26 Uh,
59:27 net outflows,
59:28 meaning that they are paying more
59:31 that service that they actually receiving financing.
59:35 In addition to all these,
59:37 there is climate change
59:40 and they need to invest in adaptation and we estimate that
59:45 it will be about 1.85% of GDP annually until 2030.
59:52 When we look at all the data,
59:54 when we look at the external financing needs.
59:58 We reached the conclusion that it will be about $820 billion
1:00:02 US dollars from 2024 to 2028.
1:00:07 And this is in a context where
1:00:09 the global financial markets remain tight
1:00:13 where we are hoping for advanced economies to start in loosening monetary policy
1:00:18 which is not happening.
1:00:19 Inflation,
1:00:20 as you know,
1:00:21 the pace of inflation has slowed down,
1:00:23 so there is lots of uncertainty about monetary policy,
1:00:27 and these countries need to raise
1:00:29 this much money.
1:00:31 So this is the context.
1:00:33 So
1:00:34 this is why DRM is so important
1:00:37 and that's why your research is so important.
1:00:40 So countries
1:00:41 will need to do a big effort.
1:00:45 On their side
1:00:46 to collect more to have better tax systems.
1:00:51 So let me tell now.
1:00:53 A bit,
1:00:54 I mean,
1:00:55 a lot of this has been said,
1:00:56 but I just want to summarize,
1:00:58 uh,
1:00:58 low income countries face many challenges one because there is
1:01:02 lots of informality.
1:01:04 Uh,
1:01:04 many countries are heavily dependent on extractive,
1:01:08 as you know,
1:01:09 many countries didn't negotiate the royalties
1:01:12 with the large oil companies in a good way,
1:01:14 so they are actually quite losers.
1:01:16 Um,
1:01:17 weak there are weak tax policy frameworks,
1:01:20 um,
1:01:21 large tax exemptions.
1:01:23 There are lots of elite capture in these countries,
1:01:26 uh,
1:01:27 and there is very limited capacity on the tax admin side.
1:01:32 Uh,
1:01:32 we heard a lot about PIT,
1:01:34 but I just want to say that
1:01:36 it has very limited role in low income countries
1:01:40 first because,
1:01:41 um,
1:01:42 the income is low across the distribution,
1:01:45 so,
1:01:46 so that is a challenge,
1:01:48 and,
1:01:49 and above all,
1:01:50 perhaps the most important point is because these economies,
1:01:53 these economies are poor.
1:01:55 So if you have a poor economy,
1:01:57 no matter how much
1:01:59 you do,
1:02:00 how well is designed is the tax system,
1:02:03 there is a limit how much you can collect.
1:02:07 Um,
1:02:07 so we are very encouraged by the research that you have done and um
1:02:13 that was presented here today and we really hope that we can collaborate and um
1:02:18 use your wisdom as we,
1:02:20 uh,
1:02:21 advise policy makers.
1:02:24 So to tell you a bit about what you are we are working and how we would like um
1:02:30 how we would like to work together and where you can help us so let
1:02:34 me,
1:02:34 let me tell you what we are doing
1:02:36 as part of the evolution roadmap DRM became front and center as I said.
1:02:42 In the past we used to do public expenditure reviews.
1:02:45 You probably have seen them.
1:02:47 You may have even contributed to some of them.
1:02:50 We used to prepare one,
1:02:52 every CPF cycle for other countries that was,
1:02:56 uh,
1:02:56 the requirement.
1:02:58 Right now
1:02:59 the requirement is that we have to
1:03:01 prepare a public finance review which will include
1:03:04 expenditure analysis and revenue analysis
1:03:09 for in every report
1:03:11 plus we need to prepare this for ID countries
1:03:15 and IBRD countries.
1:03:20 We have developed tools,
1:03:22 templates,
1:03:23 analytics,
1:03:24 and data so that the teams in the regions can
1:03:28 quickly prepare these reports because we also need to give
1:03:31 them to the ministers of financing in a timely,
1:03:33 uh,
1:03:34 timely
1:03:35 manner,
1:03:35 otherwise they are irrelevant for the policy makers.
1:03:39 Uh,
1:03:39 we are,
1:03:40 we have some pilots where we are
1:03:42 basically experimenting and see how we are doing
1:03:45 with these templates with the tools and analytics.
1:03:48 I actually encourage you all to go to the dashboard and type PFR and you'll see
1:03:54 tools,
1:03:55 analytics,
1:03:56 analytics papers,
1:03:58 data.
1:03:58 It's a very rich data set,
1:04:01 database,
1:04:02 the country partnership framework.
1:04:06 We succeeded in including a provision in policy.
1:04:11 That will require that every country that has a tax revenue to GDP below 15%
1:04:18 will need to include a discussion on the RM on the CPF,
1:04:23 as you know,
1:04:23 the CPF is the guiding document for the,
1:04:26 the,
1:04:27 for working with the government officials,
1:04:29 basically lays out
1:04:31 this country assistance strategy,
1:04:34 the assistance strategy with the client.
1:04:36 The hope is that if we discuss that in the guiding document,
1:04:41 then
1:04:42 projects and analytics will follow.
1:04:45 Pilots,
1:04:47 so we have been working on DRM for such a long time
1:04:50 with very mixed results.
1:04:52 I myself seen country reversing many policies that we decide
1:04:56 that we need to try to do things differently.
1:04:59 So we are doing this pilots with IMF where the World Bank
1:05:04 and the IMF will bring everything we have.
1:05:08 Together
1:05:10 and we will go on mission together
1:05:12 and we will have
1:05:14 joint assessments,
1:05:16 joint reports.
1:05:17 We will have
1:05:19 we will speak with one voice on what we think the countries should do
1:05:23 and we hope
1:05:25 to move the needle in a sustainable way.
1:05:28 So we have to identified 5 pilots Paraguay,
1:05:32 Cote d'Ivoire,
1:05:32 Nigeria,
1:05:33 Egypt,
1:05:34 and Bangladesh.
1:05:35 These countries will work very closely together.
1:05:38 I should say this goes beyond the RM.
1:05:40 We are also going to be working on domestic,
1:05:43 um,
1:05:44 capital mobilization.
1:05:46 So
1:05:47 we are trying to capture public money and we are
1:05:49 trying to increase public money by taxing more or better,
1:05:52 but we are also trying to mobilize private sector money.
1:05:56 Scorecard
1:05:59 If you are our president,
1:06:00 this is very important so we used to have over 200 indicators.
1:06:06 There were so many that no one really paid attention,
1:06:09 but now there will be 22 results indicator
1:06:13 that will be closely monitored
1:06:15 by senior management,
1:06:17 and one of them
1:06:19 is the RM.
1:06:21 Everyone is going to be looking into this,
1:06:23 so it's very important we count on your support.
1:06:26 And of course we provide the
1:06:29 we we write papers uh many times in collaboration with
1:06:32 you uh we do technical assistance we develop tools and uh
1:06:37 there are a number of things that we have
1:06:39 read available in the center to support the regions.
1:06:43 Um,
1:06:44 I would
1:06:45 just want to talk briefly,
1:06:46 uh,
1:06:46 about political economy even though
1:06:49 oh you able to talk about that.
1:06:51 Um,
1:06:53 tax reforms are politically very sensitive
1:06:56 actually when we had some discussions at the board
1:06:59 on the RM.
1:07:02 I've heard from some constituent constituencies that
1:07:06 it is a sovereign issue.
1:07:07 We should not talk about taxes.
1:07:10 Uh,
1:07:11 well,
1:07:11 we've been talking about taxes all the time,
1:07:13 but you know.
1:07:14 Lots of people would rather
1:07:17 uh if we would stay silent because we will uh
1:07:20 uh we will cause some noise and we may upset some people.
1:07:25 Uh,
1:07:27 The truth is
1:07:29 if
1:07:30 we do not find
1:07:32 some vested interest
1:07:34 if we don't create some noise,
1:07:37 we will never be able to move the needle
1:07:39 and have impact on the RM.
1:07:42 And there are some encouraging cases.
1:07:44 So for example,
1:07:46 Kenya,
1:07:46 they were about to have,
1:07:47 uh,
1:07:48 a liquidity cliff,
1:07:49 and they managed to avoid it because they started to implement some reforms.
1:07:53 Angola
1:07:54 implemented the full subsidy reform.
1:07:56 Ghana,
1:07:57 as you know,
1:07:57 they had the
1:07:58 domestic debt restructuring.
1:08:00 They are in the process of external debt
1:08:02 restructuring and they are implementing reforms already.
1:08:05 Nigeria,
1:08:07 they
1:08:07 eliminated full subsidies.
1:08:10 They came back.
1:08:11 It's a bit of a,
1:08:13 well,
1:08:13 they go,
1:08:14 come back,
1:08:14 they go back,
1:08:15 they come back,
1:08:15 but anyway,
1:08:16 it's a bit difficult,
1:08:17 but they have also done some DRM reforms.
1:08:20 I didn't put here,
1:08:21 but Egypt is another case.
1:08:24 But
1:08:24 I wanted to um.
1:08:28 Say,
1:08:28 uh,
1:08:29 two more things on this point.
1:08:31 One is that we are talking a lot about the RM,
1:08:36 but Yel showed an important fact when you said that advanced economies
1:08:40 were
1:08:41 collecting as much as
1:08:43 the low income economies are
1:08:45 collecting now 100 years ago.
1:08:47 What happened in between these 100 years?
1:08:49 They grew.
1:08:51 So we can
1:08:52 think about the most fantastic reforms if the economies do not grow,
1:08:56 we will not achieve results
1:08:58 so we cannot neglect
1:09:00 growth reforms
1:09:02 to complement tax reforms.
1:09:04 And finally,
1:09:05 um,
1:09:06 when we think about development,
1:09:08 we cannot,
1:09:09 in my view we can never focus on revenues only
1:09:13 because
1:09:13 how you spend it
1:09:14 matters as much as how much you
1:09:17 resources you collect.
1:09:19 If you,
1:09:19 you collect the resources and they go into waste
1:09:22 into some white elephants in terms of infrastructure,
1:09:26 if they don't go into productive use,
1:09:28 if they are not gonna full growth of the future.
1:09:31 In the end you will not achieve your objectives,
1:09:34 so it's very important to always look on the expenditure side too.
1:09:39 Um,
1:09:39 so in our team,
1:09:40 very briefly.
1:09:42 We are working on uh
1:09:45 cost effectiveness of tax incentives,
1:09:47 particularly in VAT.
1:09:48 We are broadening the base of the VAT systems.
1:09:52 We are doing some reforms on PIT,
1:09:54 but we are also looking to personal income taxes
1:09:57 if we want to talk.
1:09:58 About progressivity typically
1:10:00 uh income taxes are
1:10:02 paid by the,
1:10:03 the wealthiest so even if you change on the margin is gonna be progressive
1:10:08 and property taxes which are very important,
1:10:11 um,
1:10:11 as well,
1:10:12 but you need the number of forms,
1:10:14 um,
1:10:15 namely fiscal cadus,
1:10:16 uh,
1:10:17 violation systems
1:10:18 actually and talked about Pakistan a lot.
1:10:21 I worked on Pakistan.
1:10:23 And actually we started some reforms in the area of
1:10:27 of fiscal cadast evaluation.
1:10:31 So
1:10:32 So it was not in all province,
1:10:34 but in some provinces
1:10:36 the land was not they were not collecting taxes on
1:10:39 on land,
1:10:41 and the result was that is that the land stayed
1:10:45 idle.
1:10:46 No one was using that land
1:10:48 for a productive use because there were no opportunity cost on that land.
1:10:53 So there are two problems
1:10:55 is that one,
1:10:55 you do not collect and typically these
1:10:58 are to support,
1:10:59 uh,
1:10:59 local government.
1:11:00 Second.
1:11:01 There is no opportunity cost,
1:11:03 so you just leave.
1:11:05 Uh,
1:11:05 you just leave land
1:11:07 without any use and without,
1:11:09 uh,
1:11:09 making,
1:11:10 um,
1:11:10 a productive use of,
1:11:12 of it,
1:11:13 and we are also working on tax administration reform
1:11:17 that is an agenda more for our colleagues,
1:11:19 uh,
1:11:19 in governance,
1:11:21 um,
1:11:21 so I just want to like to conclude with the collaboration
1:11:26 with DC,
1:11:27 uh,
1:11:28 so we are already doing some work together,
1:11:31 notably in Nigeria potentially Zimbabwe and others.
1:11:35 Um,
1:11:36 the microdata that we heard about today I think can
1:11:39 be quite helpful for us as we support the regions.
1:11:43 Uh,
1:11:43 the data lab is something we are very proud of,
1:11:47 and,
1:11:48 uh,
1:11:48 and it,
1:11:49 it's a joint effort
1:11:50 and we think we can really,
1:11:52 um,
1:11:53 take that far
1:11:54 and I think can be very,
1:11:55 very helpful as we continue our support to the countries
1:11:59 and,
1:11:59 uh,
1:11:59 yes,
1:12:00 with this,
1:12:00 let me conclude.
1:12:01 Thank you so much.
1:12:08 Thanks Manuela.
1:12:09 That was amazing that even the discussant ended with the day tax plug.
1:12:14 That's pretty wild,
1:12:16 um,
1:12:17 the,
1:12:18 well,
1:12:18 thank you.
1:12:19 That was that was great reflections,
1:12:21 um,
1:12:22 on the content,
1:12:23 on the collaboration,
1:12:25 on the synergies between the,
1:12:26 the research and the operations,
1:12:29 um.
1:12:31 I,
1:12:32 uh,
1:12:32 just on the issue of,
1:12:34 of land actually,
1:12:35 just to give another plug which is at the land conference that's coming up very soon,
1:12:40 we'll be doing a launch of a report that the research
1:12:43 group we're doing in collaboration with the Africa Chief Economist Office.
1:12:46 Uh,
1:12:46 where this issue of land,
1:12:48 land registration and linked to taxation,
1:12:51 uh,
1:12:52 is,
1:12:52 is,
1:12:52 uh,
1:12:54 highlighted and a really salient aspect of,
1:12:56 of land policy
1:12:58 in sub-Saharan Africa,
1:13:00 uh,
1:13:00 uh,
1:13:00 one of the thing I wanted to compliment you on
1:13:02 was
1:13:03 not only did you the reflections really useful,
1:13:07 uh,
1:13:07 uh,
1:13:07 was the.
1:13:08 You actually made the evolution roadmap sound
1:13:11 fairly clear and straightforward.
1:13:15 That's kind of the first time I've seen that.
1:13:18 Um,
1:13:18 OK,
1:13:19 so,
1:13:19 uh,
1:13:19 without further ado,
1:13:20 let's,
1:13:21 well,
1:13:21 for the presenters,
1:13:22 are there any immediate reactions you might have to,
1:13:25 to,
1:13:25 to the discussion,
1:13:27 um,
1:13:27 and,
1:13:28 uh,
1:13:29 otherwise we can just open it up for,
1:13:31 for,
1:13:31 for general questions.
1:13:33 Um,
1:13:34 if you're online again,
1:13:35 just raise your hand or put,
1:13:37 put something in the chat that you'd like to ask a question.
1:13:40 Maybe we'll start from within the room,
1:13:41 just raise your hand and I get a sense of how many we have one Norman there.
1:13:46 For now we just have one.
1:13:47 OK,
1:13:47 Norman,
1:13:47 go ahead.
1:13:51 Thank you.
1:13:52 Uh,
1:13:52 well,
1:13:53 first of all,
1:13:53 congratulations.
1:13:54 I think it's wonderful research
1:13:57 and uh very well presented.
1:13:59 I do have a challenge for for you guys,
1:14:02 and this is to write
1:14:03 another set of papers
1:14:05 where the emphasis is taxation for growth.
1:14:09 Cause we have seen lots of interesting presentations here on how to tax more.
1:14:15 Um,
1:14:16 how to make the VAT more efficient in the same regard
1:14:20 and also taxation for equity.
1:14:23 But there is this other aspect
1:14:24 that
1:14:26 I believe is at least as important,
1:14:29 and this is
1:14:30 taxation so that you can increase the base you can
1:14:33 have
1:14:34 less informality
1:14:36 and you can actually have
1:14:38 more growth
1:14:39 and you know all the benefits that come with
1:14:42 growth,
1:14:42 of course,
1:14:43 and one of those benefits might actually be
1:14:46 you will be able to collect more tax revenues.
1:14:49 So you can then
1:14:50 you can then complete the circle.
1:14:53 Thanks.
1:14:57 Thanks,
1:14:58 Norman.
1:14:58 Anybody else in the room with a question
1:15:01 or reflection?
1:15:06 Oh,
1:15:06 there,
1:15:07 go ahead.
1:15:09 So maybe this is an opportunity to make a plug
1:15:12 um I think the one takeaway I have on
1:15:15 the presentations and then I'll link it to the question
1:15:18 that was just asked is that
1:15:20 traditionally it's often assumed that there is a
1:15:23 trade off between efficiency and equity
1:15:27 and I think a very important message that's coming out from the analysis of tax
1:15:32 administration and other data
1:15:34 that Pierre
1:15:35 Anne Oyebola and other colleagues have been making is that.
1:15:39 This is not always the case.
1:15:41 In fact,
1:15:42 efficiency improvements
1:15:44 can be progressive.
1:15:46 So for example,
1:15:47 the,
1:15:47 the,
1:15:47 the mention,
1:15:48 the point that they make on VAT exemptions
1:15:51 is that we find that most VAT exemptions are captured.
1:15:54 Actually by the larger businesses and benefits,
1:15:58 the higher income deciles
1:15:59 means by definition that if you take this away
1:16:02 and you use this VAT,
1:16:04 even in a universal cash transfer,
1:16:05 which is another paper,
1:16:07 we find the result is more progressivity.
1:16:09 So you're both improving efficiency
1:16:11 and you're improving equity at the same time.
1:16:13 Supporting growth
1:16:14 and you're supporting the inclusivity of this growth
1:16:17 so that's um I think that that's an important takeaway for me that when I see this work
1:16:22 and the the plug is we're attempting to start to get into
1:16:26 this space but we fully agree that this is a knowledge gap
1:16:29 we have a PFR pilot that Manuela mentioned is Georgia.
1:16:32 It's going for decision review tomorrow morning,
1:16:36 so please don't kill us,
1:16:37 but in that tax chapter
1:16:39 we look at informality,
1:16:40 tax efficiency,
1:16:42 and growth,
1:16:42 and we have some policy recommendations on how
1:16:45 you can basically change the tax mix
1:16:48 and reform some
1:16:49 tax exemptions
1:16:50 to basically reduce informality,
1:16:52 and we show with a with a
1:16:54 CG model that that will boost productivity and growth.
1:16:57 So have a look at that,
1:16:59 but that's only the tip of the iceberg.
1:17:00 We need to do a lot more.
1:17:03 Thanks.
1:17:04 I want to give a chance to the presenters or or or or or the discussant
1:17:08 if they want to react to these comments.
1:17:10 Do you have a,
1:17:11 OK,
1:17:11 go ahead.
1:17:13 Thank you Norman.
1:17:13 I think it's a good point,
1:17:14 but I,
1:17:15 I agree with what Jafar said to some extent,
1:17:17 right?
1:17:17 And I tried to say it in the last slide,
1:17:19 which is
1:17:20 it's not obvious these days that there's always
1:17:22 so much tension between the goal of efficiency and equity.
1:17:25 In the sense that you know
1:17:27 from the work we've done with that on effective tax rate of firms,
1:17:29 you observe that even within the very large firms
1:17:32 you have huge gaps in the effective tax rate.
1:17:34 There's a lot of firms who are literally paying 0%
1:17:36 and then there's firms who are abiding,
1:17:37 you know,
1:17:38 by a 30% tax rate,
1:17:39 for example.
1:17:40 It's really hard to think,
1:17:41 you know,
1:17:41 in which world this is an efficient policy to have,
1:17:43 you know,
1:17:44 firm specific tax rates basically,
1:17:46 which is what has happened.
1:17:47 A lot of Latin American countries for example
1:17:49 via you know special economic zones that were supposed
1:17:52 to be for exporters but when you look a lot of these firms are not exporting anything
1:17:56 and the likes right so.
1:17:58 You know,
1:17:59 currently I think we really need to be thinking
1:18:00 of closing a lot of the loopholes which you know
1:18:03 are inefficient they lead to a lot of rent seeking,
1:18:05 you know,
1:18:05 optimization I mentioned,
1:18:06 you know,
1:18:07 you transform your capital income into labor income and vice versa,
1:18:10 you know,
1:18:10 a lot of things that actually costly for the firms.
1:18:13 In terms of accounting,
1:18:16 um,
1:18:17 and then you know if you've done that,
1:18:19 then you can think of how do you,
1:18:21 sorry,
1:18:21 so
1:18:22 could you please close your microphone online?
1:18:26 I,
1:18:27 the person who's speaking,
1:18:28 can you please,
1:18:29 uh.
1:18:31 Ale Alexandra,
1:18:31 could you please put mute?
1:18:32 Thank you,
1:18:34 thank you,
1:18:34 uh,
1:18:35 and you know if you have achieved that then there
1:18:36 is a question how do you recycle the revenue?
1:18:39 Can you even lower some of the tax rates?
1:18:41 The real risk is that if you don't do that
1:18:43 you're gonna have to increase taxes on payroll,
1:18:44 for example,
1:18:45 or on consumption,
1:18:46 right,
1:18:47 which
1:18:47 they're also very distortive
1:18:49 and on top of that they hit more,
1:18:50 you know,
1:18:50 the median earner or even poor earners,
1:18:52 so.
1:18:54 Sorry,
1:18:54 quick get 200,
1:18:56 uh,
1:18:56 Pierre,
1:18:56 I hope that you're not dismissing
1:18:58 the research agenda on,
1:19:00 uh,
1:19:00 taxation and growth
1:19:02 because what you just said is that with one slide.
1:19:05 You're you're summarizing whatever we need to know
1:19:08 about taxation and growth,
1:19:09 and I don't think that's true.
1:19:11 No,
1:19:11 no,
1:19:11 and I agree with you,
1:19:12 you know,
1:19:12 we can think of a lot of,
1:19:14 you know,
1:19:14 policies,
1:19:15 but I think right now we're starting from way
1:19:17 too many incentives that are very badly targeted,
1:19:19 right?
1:19:20 And so that doesn't mean there's not a big role of taxes for growth,
1:19:23 but I think thinking of this
1:19:25 matters a lot,
1:19:26 right?
1:19:26 And so that can give you almost a double dividend I think.
1:19:29 Once you've plucked some of these,
1:19:30 I agree with you,
1:19:31 you know,
1:19:31 and I think when we're thinking of the transition right
1:19:34 towards a green,
1:19:35 greener economy,
1:19:36 we're gonna need to think a lot of,
1:19:37 you know,
1:19:38 how the tax system works and so on.
1:19:39 So I agree with you,
1:19:40 Norman.
1:19:40 Sorry,
1:19:41 didn't mean that,
1:19:41 but I,
1:19:42 I'm not sure that,
1:19:42 you know,
1:19:43 a lot of people always put these two efficiency and equity in contradiction.
1:19:46 I'm not sure that's necessarily always the case,
1:19:48 right?
1:19:50 So,
1:19:51 Oyebola and Anne,
1:19:52 any,
1:19:52 any other reactions?
1:19:55 So.
1:19:56 OK,
1:19:56 so I have a couple of questions.
1:19:59 Um,
1:20:00 so Oyebola,
1:20:00 I want,
1:20:01 I wanted to come back to the
1:20:04 what I thought was a,
1:20:05 a,
1:20:05 a,
1:20:06 an odd contrast,
1:20:07 but then,
1:20:07 but then Manole kind of picked up on it as well,
1:20:10 which is the
1:20:11 sort of historical trajectory of tax and GDP rates.
1:20:15 Which immediately,
1:20:16 I mean,
1:20:17 you anticipated the way we would interpret that is,
1:20:19 oh,
1:20:19 as countries grew,
1:20:21 as incomes get higher,
1:20:22 then,
1:20:23 you know,
1:20:24 the,
1:20:24 the,
1:20:24 the,
1:20:24 the tax to GDP rate goes up.
1:20:27 So countries today
1:20:29 are just like
1:20:30 these other countries were 100 years ago.
1:20:33 But then the next slide,
1:20:33 you showed that there's basically no association
1:20:37 between GDP per capita
1:20:39 and or I guess it was GDP per capita growth,
1:20:41 OK.
1:20:42 So
1:20:43 anyway,
1:20:44 so is,
1:20:45 so maybe the the simple version of the question,
1:20:47 is it true that low income countries today are just like
1:20:51 the,
1:20:52 the rich countries today were when they were low income or middle income?
1:20:56 Is that statement sort of
1:20:57 about right?
1:20:58 Yes,
1:20:59 so I think the.
1:21:00 Income levels
1:21:02 they're very similar in that in that regard um
1:21:05 the
1:21:06 main takeaway from this is that I completely agree with um Manuela's point about
1:21:11 growth is essential.
1:21:12 That's like you have to start from there,
1:21:14 but it's not automatic
1:21:16 that once you have higher incomes then your
1:21:18 tax collections will just automatically also take off
1:21:21 because you'll find many places where
1:21:23 there's economic growth but
1:21:25 they aren't this complimentary investments and actually.
1:21:27 Harnessing this growth and converting that into
1:21:31 the revenue collection which can then be reinvested and so
1:21:34 um
1:21:35 I think that's the main point there is that you
1:21:38 can have the economic growth without
1:21:40 actually increasing your
1:21:42 um revenues and
1:21:43 we also find reversals where some countries
1:21:46 actually were collecting
1:21:48 you know um
1:21:49 I was looking reviewing something for Indonesia
1:21:51 recently they were collecting pretty high.
1:21:54 You know,
1:21:54 like a decade,
1:21:55 two decades ago and now really dropped.
1:21:57 And so they're all,
1:21:58 there are lots of things that happened,
1:22:00 which
1:22:02 can be quite parallel to the growth trajectory
1:22:04 of the,
1:22:05 of the country.
1:22:08 I guess I had not,
1:22:09 thank you.
1:22:10 That's clarifying,
1:22:11 um,
1:22:12 for,
1:22:12 for Anne,
1:22:13 I,
1:22:13 I hope you're still there.
1:22:14 You,
1:22:14 you,
1:22:15 yes,
1:22:15 there you are.
1:22:16 I guess
1:22:17 one of the phrasings you used was that alternatives are not viable,
1:22:22 and
1:22:22 I guess I just want to come back to that term viable
1:22:25 in a sense I thought what you showed us is that maybe they're not
1:22:28 as effective or they,
1:22:30 they're not as efficient or they're more expensive.
1:22:33 But I,
1:22:33 I'm kind of curious why you use the term viable.
1:22:35 Is that they're not sustainable or what,
1:22:37 what,
1:22:38 how,
1:22:38 how are you thinking?
1:22:40 What were you thinking of when you use the term viable?
1:22:45 It's,
1:22:45 it's a good question.
1:22:46 Um,
1:22:47 yeah,
1:22:47 so I guess I was thinking of viable
1:22:49 within the current budgetary framework of those countries
1:22:54 because it seems that the retail sales tax
1:22:57 would make the current kind of fiscal expenditure predictions of
1:23:01 the countries not viable because it would just collect.
1:23:03 So much less tax
1:23:05 and the turnover tax similarly would probably
1:23:09 collect much less revenue and also be very distortionary to
1:23:12 growth so um to to Norman's challenge then I think the
1:23:16 turnover taxes
1:23:17 yeah
1:23:18 it's definitely something to to that countries need to to stay away from.
1:23:23 Um,
1:23:24 when they want to,
1:23:25 yeah,
1:23:25 have
1:23:26 both,
1:23:27 uh,
1:23:27 um,
1:23:28 strong governments that can ensure pro-poor expenditures,
1:23:32 uh,
1:23:32 but also support,
1:23:33 uh,
1:23:34 the development and growth of the economy,
1:23:35 uh,
1:23:36 more generally,
1:23:37 um,
1:23:37 and.
1:23:38 Maybe just to uh add a comment on Norman's challenge which is uh a good one.
1:23:43 So I think in the Datex program we would have a whole work stream
1:23:47 on production efficiency which is basically about
1:23:49 kind of relieving constraints to growth,
1:23:52 um,
1:23:53 constraints to resource allocation.
1:23:55 Um,
1:23:56 thinking about how tax policy,
1:23:58 whether and how tax policies should,
1:24:00 uh,
1:24:00 support the green transition would also hopefully become part of that,
1:24:04 um,
1:24:05 because with,
1:24:05 uh,
1:24:05 tax incentives for the green transition kind of proliferating,
1:24:08 there is kind of the,
1:24:09 the race.
1:24:10 To the bottom in green tax incentives or the race in green tax incentives,
1:24:15 um,
1:24:15 and I also just wanna mention that there's already quite a bit of
1:24:18 ongoing work on formalization and design of tax systems for small firms.
1:24:22 So for instance Chris Hoy,
1:24:23 Jonathan Carver,
1:24:24 and Tiago Scott are working on that in,
1:24:26 uh,
1:24:26 Georgia,
1:24:27 Kenya and Tanzania also connected with the PFL work,
1:24:30 uh,
1:24:30 that Jaafar was mentioning.
1:24:32 So we hope soon also to have new insights on how to
1:24:34 encourage the transition of these small firms into the formal sector.
1:24:40 Thanks,
1:24:40 actually,
1:24:41 that relates to uh a last question I had for,
1:24:43 for Pierre,
1:24:43 which is
1:24:44 sort of this notion of informality and the way you present,
1:24:46 you almost presented it like um.
1:24:49 Uh,
1:24:49 like it was,
1:24:50 uh,
1:24:51 either an explicit or an explicit exemption,
1:24:54 and you know,
1:24:55 you,
1:24:55 you sort of incorporate that into your assessment of progressivity,
1:24:58 but is that,
1:24:58 is that really sustainable?
1:25:00 I mean,
1:25:01 sure,
1:25:01 as a sort of a snapshot one time,
1:25:03 OK,
1:25:04 but
1:25:07 should a country be thinking of that as a policy per se that that they want to actually
1:25:11 treat explicitly?
1:25:13 Thanks Dion.
1:25:14 That's a that's a very good question.
1:25:15 So
1:25:16 first I think coming back indeed,
1:25:18 you know,
1:25:18 without growth and
1:25:20 well targeted expenditure,
1:25:22 you know,
1:25:23 like a lot of what we've been talking about is not that interesting,
1:25:25 right,
1:25:26 in the sense that you don't probably wanna collect
1:25:28 much more revenue if it's all wasted and.
1:25:30 In a context
1:25:32 where you know the country is going backwards in a way right?
1:25:33 so
1:25:34 that I think we all agree upon maybe we didn't make it clear enough
1:25:37 and so I think that's gonna link to your question in the sense that
1:25:40 should we think of the informal sector it's size of something structural
1:25:44 that there's not a whole lot you can do about it or is that a parameter in and of itself
1:25:48 increasingly and I think you know some of the research for example by David McKinsey
1:25:51 and others have shown that it's not easy you know just with small incentives,
1:25:56 you know,
1:25:57 and so on to register a lot of firms.
1:25:59 And not is it obvious that you actually wanna do so what you really want is firms
1:26:03 to select basically into formality because there's some
1:26:05 advantage eventually when they become big enough for example
1:26:08 you get better credit access,
1:26:10 you know,
1:26:10 you get
1:26:11 uh potentially other advantages.
1:26:13 Uh,
1:26:14 and so that's probably,
1:26:15 you know,
1:26:15 the link to growth again should be made here is that as
1:26:17 you get bigger you're gonna wanna select anyways into the formal sector
1:26:22 at least that's my view and my reading of the literature,
1:26:23 you know,
1:26:24 when you think of some of the small,
1:26:25 smaller firms you could do some
1:26:27 specific intervention it's probably not gonna be very long lasting and long lived
1:26:31 and so my view is to see this as something
1:26:33 a little bit more structural we can play with it,
1:26:36 you know,
1:26:36 at the margin.
1:26:37 Uh,
1:26:38 but I,
1:26:38 I tend to think that,
1:26:39 you know,
1:26:40 that
1:26:40 if there's a lot of traditional or self pro
1:26:42 shops or self-production that is quite structural to the economy
1:26:45 and there's not a whole lot maybe policy can do.
1:26:49 OK,
1:26:49 great.
1:26:49 Well,
1:26:50 we've we've come to the end of our time.
1:26:52 Did you have any closing reactions you'd like to
1:26:55 have just on this point of taxation for growth.
1:27:00 Uh,
1:27:01 for me everything is for growth,
1:27:03 so
1:27:04 taxation should be for growth.
1:27:06 Expenditure policy should be towards promoting growth.
1:27:09 Debt policy should be towards sustainable growth.
1:27:12 So,
1:27:13 um,
1:27:15 so if it's,
1:27:16 if not for growth,
1:27:17 for what,
1:27:18 right?
1:27:20 No,
1:27:20 but in my view,
1:27:22 if you want to have equity in the distribution,
1:27:24 you need to grow.
1:27:26 The for me growth is not in the trade offs.
1:27:28 There are trade offs.
1:27:30 There might be trade offs,
1:27:31 but if you don't grow,
1:27:32 you don't achieve much,
1:27:33 so that is the base.
1:27:34 So everything is to promote growth.
1:27:37 So,
1:27:37 um,
1:27:39 I,
1:27:39 I,
1:27:40 what I really think it's,
1:27:42 it's,
1:27:42 it's important to design policies
1:27:45 that support economic growth the same way expenditure as I said,
1:27:49 but at times.
1:27:50 There is a bit of a confusion when you speak about taxation for growth.
1:27:55 People may interpret in in some countries
1:27:58 this could be interpreted as tax incentives
1:28:02 that would allow
1:28:04 to attract some companies into a country
1:28:07 because of tax breaks,
1:28:08 tax credits,
1:28:09 or tax exemptions
1:28:11 at times creates that confusion and we don't want that confusion.
1:28:15 That's
1:28:18 really good the research.
1:28:21 OK,
1:28:22 well,
1:28:22 thank you everybody.
1:28:23 Please join me in thanking the presenters,
1:28:24 our discussion for a really fascinating.
- add-style
- lp-body-content