Government procurement is big business, typically accounting for 10 to 20 percent of annual GDP depending on the country and year. Given this scale and the high level of discretion involved, it is hardly surprising that policy makers often view procurement systems as a powerful industrial policy tool to achieve macroeconomic policy objectives.
In this Policy Research Talk delivered on January 28, 2025, World Bank Senior Economist Manuel GarcÃa-Santana presented findings on the impacts of policies such as "buy small" and "buy local" on the private sector and the broader economy. His talk provided participants with a more nuanced understanding of the trade-offs involved in designing procurement policies.
00:16 OK,
00:17 let's get going.
00:18 Welcome,
00:18 everybody.
00:19 Um,
00:21 Thank you for coming.
00:22 I guess the people outside will maybe trickle in.
00:24 If somebody could just wave to them and encourage them.
00:28 Um,
00:28 thank you.
00:29 Um,
00:33 So I'd like to welcome you all to today's policy research talk.
00:37 um,
00:38 these talks,
00:38 uh,
00:39 which we try to schedule about,
00:40 uh,
00:41 once a month,
00:41 um,
00:42 give us an opportunity to present work coming out of the research department
00:46 with
00:46 each other,
00:47 with colleagues in the rest of the World Bank,
00:50 as well as,
00:51 um,
00:51 many people outside the World Bank.
00:53 Um,
00:54 so I'd like to welcome.
00:55 All of you in the room and those of you online watching the Webex and the live stream.
01:02 Um,
01:02 today we have a topic of particular importance,
01:06 which
01:06 we're just chatting about outside.
01:08 We don't actually do a lot of research on the research department.
01:11 So this is a new kind of topic for us,
01:13 which is,
01:14 uh,
01:14 government procurement.
01:15 Uh,
01:16 public procurement accounts for 10 to 20% of annual GDP,
01:20 um,
01:21 you know,
01:22 depending on the country and the year,
01:23 um,
01:24 making it a substantial economic driver,
01:28 and
01:29 particularly in some of the world's poorest countries.
01:32 Um,
01:33 with the scale comes a lot of potential and potential for
01:38 using procurement to achieve different goals,
01:40 such as policies like buy small,
01:42 buy local,
01:43 buy green.
01:44 And these are often seen as important
01:46 industrial policy tools for advancing macroeconomic objectives.
01:50 Um,
01:51 but I guess,
01:52 you know,
01:53 lots of questions arise.
01:54 Are,
01:54 are,
01:55 are these effective?
01:56 Do they lead to unintended consequences?
01:59 What are the macro,
02:00 you know,
02:00 what,
02:01 what's the size of the macroeconomic impacts?
02:04 And also,
02:05 you know,
02:06 understanding how this plays out across different
02:08 countries at different levels of development.
02:11 So to help us explore these questions,
02:13 uh,
02:13 we're joined by our presenter today,
02:14 Manuel Garcia Santana.
02:16 He's a senior economist in the macroeconomics and
02:19 growth unit of the Development Research Group.
02:22 Manuel's research focuses on the macro aspects of development and trade.
02:27 Um,
02:27 before joining the bank,
02:28 he served as an associate professor at Pompeo-Fabbra University
02:32 and was a Canon Fellow at
02:33 Princeton University's International Economics section.
02:38 Uh,
02:38 I'm also delighted to welcome our discussant,
02:40 Doctor Majid Elbaya,
02:42 the regional procurement manager,
02:45 uh,
02:45 for the Middle East and North Africa region at the World Bank.
02:48 He's joining us from Djibouti today.
02:50 And would you mind putting us,
02:51 uh,
02:52 put the,
02:52 the sort of participants on the screen just so we can see
02:55 him and welcome,
02:56 uh,
02:57 welcome.
02:58 Thank you for joining us,
02:59 uh,
03:00 from so far and so late.
03:02 Um,
03:03 Doctor Ebeyat brings an extensive experience in public procurement reforms,
03:08 country reviews,
03:09 fiduciary assessments,
03:10 and capacity building initiatives.
03:12 He has also led and contributed to several studies on public
03:15 procurement.
03:16 He holds a PhD and master's degree in civil engineering,
03:20 hydraulics
03:21 from the University of Missouri Columbia,
03:23 and a BSC in civil engineering from,
03:26 uh,
03:27 Bit University in Palestine.
03:29 So the format for today is Emmanuel will present for about 45 minutes,
03:34 um,
03:34 after which we'll hear from Majed for about 10 minutes or however long he'd like to,
03:38 to intervene.
03:40 Then we'll open up the floor for Q&A,
03:42 uh,
03:42 from the audience.
03:42 If you're online,
03:43 please raise your hand or,
03:45 you know,
03:46 put the,
03:47 that you have a question in the chat and we'll,
03:49 we'll hopefully
03:50 be able to call on you and if you could,
03:52 wouldn't mind,
03:53 uh,
03:53 unmuting and,
03:55 um,
03:55 and actually intervening.
03:57 Uh,
03:58 could you mute if you're,
04:00 uh,
04:01 participating?
04:02 Thank you.
04:03 Um,
04:04 with that,
04:05 uh,
04:05 over to you,
04:05 Manuel.
04:12 Can you hear me?
04:12 OK,
04:13 thank you,
04:13 Dion for the introduction and thank you all for being here today.
04:24 Just waiting for the slide
04:26 to show up.
04:28 OK
04:30 Um,
04:35 Sure this show the control pad.
04:46 It's this one.
04:46 No.
04:49 Yeah,
04:49 there you go.
04:57 People
04:58 cannot see it there.
05:01 You update the uh
05:30 Thank you.
05:32 OK,
05:33 so why is
05:34 um government procurement important?
05:36 So Dion already went through some of my introduction,
05:38 so that was very helpful.
05:40 uh I'm just gonna
05:42 be repetitive.
05:44 So
05:45 let me first define what government procurement is so
05:47 the way you can think about it is governments
05:50 around the world buy
05:52 lots of goods and services from private firms,
05:55 so the cost of all the purchases is
05:58 government procurement.
06:00 So feature number one,
06:02 as Dion emphasized,
06:03 is huge,
06:04 accounts for
06:05 10-20% of GDP depending on the country and year.
06:08 Feature number two
06:10 is spread across the whole economy,
06:11 so governments buy
06:13 almost everything.
06:14 They buy from
06:15 printing paper to IT services.
06:17 They build bridges.
06:18 They buy,
06:19 uh,
06:19 toilet paper,
06:20 they buy books,
06:21 they buy chairs.
06:21 They buy basically everything.
06:25 Feature number 3,
06:26 there are many firms that are involved in government procurement.
06:29 So as measured by the enterprise survey,
06:32 around 15% of firms on average
06:35 report that they were trying to compete for at least one government contract
06:39 in the preceding year.
06:41 So a lot of firms matter for government procurement.
06:45 Then it's often perceived as a powerful policy tool so it's
06:49 especially in the last few years public procurement
06:51 has been at the core of the policy debate
06:53 and has been emphasized as a tool governments can use to affect
06:56 the economic outcomes for example in the case of the EU.
07:00 Has been at the core of the industrial
07:01 policy reports by the European Commission emphasizing that making
07:05 a public procurement market better is gonna affect the economy
07:09 uh positively in many dimensions.
07:12 In the context of the ban,
07:13 it has been emphasized that reducing the the the waste in procurement
07:18 may allow governments to use these extra resources to do more with this.
07:23 And very recently in the context of the US
07:26 buy American,
07:27 which is a policy that I'm gonna provide much more details,
07:30 but it's a policy that
07:32 imposes restrictions on the type of products and services that federal agencies
07:36 can buy in the US has been at the really core of how
07:39 the federal government may actually help to create the uh jobs
07:43 and and economic activity in the manufacturing sector in the US.
07:48 Number 5,
07:49 not only there seems to be kind of
07:51 an importance of procurement in levels,
07:53 but there seems to be a trend
07:55 that is gonna make public procurement even more important over time.
07:59 So this is inspired by the recent paper by Juhas and
08:02 co-authors where they use this data set Global Trade Alarm,
08:05 which you can measure
08:06 all types of policies around the world.
08:09 So what I did is that using a predetermined definition they have in this data set
08:14 which is about procurement procurement related policy
08:17 and what I did was plotting the
08:19 number of procurement related policies that were announced
08:23 across the different years
08:25 and what you can see is that there were like some kind of constant trend for 10 years
08:30 and starting in 2017 there was kind of an upward trend in the
08:34 number of procurement related policies that were
08:36 announced by countries around the world.
08:39 Well,
08:39 you may think that this trend
08:41 is simply driven mechanically by the fact there is more regulation over time
08:45 in all aspects of policy.
08:47 So in the second graph what I'm plotting is the procurement
08:50 related policies as a fraction of all the policies that governments announce
08:55 over time.
08:56 And what you can see is that nowadays procurement related policies account for
08:59 around 6% of total policies around the world as measured by these data.
09:03 That's a huge number because by all policies I mean.
09:06 Everything that has to do with anti-dumping trade rules of origin,
09:09 credit subsidies,
09:10 and so on and so forth.
09:12 So summary procurement is huge spread across the whole economy
09:16 is important for private firms.
09:19 There seems to be at the core of the policy debate and
09:21 it's important is increasing over time in terms of policy reforms.
09:25 So what is gonna be the plan for the talk?
09:27 The plan for the talk is gonna be the
09:28 following first I'm gonna give you concrete examples of
09:32 procurement policies that are
09:34 very important in trying to understand the
09:35 type of mechanisms I will emphasize later.
09:38 The second part of the talk is gonna be
09:41 fixing what is the thought experiment that I want
09:43 you to keep in your mind throughout my presentation.
09:45 What is the type of counterfactual we want to have in mind
09:48 that is gonna be helpful for policymakers?
09:51 Third part is gonna be
09:53 providing you with a framework that I think is general
09:57 to actually trying to come back to the thought
09:58 experiment and come up with answers and mechanisms.
10:02 Then I'm gonna focus on two particular applications based on
10:05 my research with some co-authors and I'm gonna just conclude.
10:09 Examples of procurement policies,
10:10 so buy America,
10:12 this started uh
10:14 you know,
10:15 in the 30s,
10:16 but that has some that I mean it has been super,
10:18 super important in the last few years.
10:21 Uh,
10:21 in the US administration,
10:23 so how does it work?
10:24 The way it works is that federal agencies have restrictions on the
10:28 choices they make regarding the suppliers they choose and in particular,
10:32 for example,
10:32 the 20% bid margin means that unless the price that is offered by
10:37 the foreign firm is 20% lower than that offered by the domestic firm,
10:42 then the federal agency has to buy from the domestic firm.
10:45 So this is a way in which governments are explicitly
10:47 telling you how much they're willing to pay extra.
10:50 In order to buy from a local firm.
10:53 And this is very prevalent around the world.
10:55 Buy
10:56 less local or domestic this is kind of the opposite
10:59 of the buy American in the case of the EU,
11:02 uh,
11:02 the European Commission has been trying to unify the market for procurement in a
11:06 way that somehow resembles the single market spirit that the EU had in mind.
11:11 So
11:12 by harmonizing rules and making sure that governments implement an
11:15 equal treatment across firms from all all uh origins,
11:19 the EU was thinking about,
11:21 you know,
11:21 increasing market integration and potentially improve the
11:24 procurement outcomes at the EU level.
11:27 Buy small is also very prevalent,
11:29 so there are reservations for a small firms so governments impose themselves
11:33 some restrictions just to make sure
11:35 that a particular fraction of total expenditure
11:37 is actually allocated to small firms and again
11:39 this is very prevalent around the world.
11:42 There are some other policies that are gonna
11:44 have less importance in my presentation but I think
11:47 are gonna gain a lot of importance in the coming years like for example the buy green
11:51 so there are already governments that are using
11:53 their procurement systems to make sure that.
11:55 The economy becomes greener by buying from greener firms or greener products
11:59 also buy from minorities.
12:01 There are many governments that think that this government demand
12:04 can be used to help minorities that may be constrained somehow
12:07 to help them to overcome these constraints
12:09 and also for sure reduce waste and corruption,
12:12 which has been kind of the classical view
12:14 of public procurement.
12:16 So what is the thought experiment I'm gonna be thinking throughout my
12:19 presentation so I really want you to fix that in your mind.
12:22 Imagine
12:23 the government changes the composition of firms from which it buys.
12:27 So remember I told you governments buy toilet paper,
12:29 they buy printing paper,
12:30 they buy IT services.
12:31 Imagine they just keep buying the same stuff
12:33 they just change the firms from which they buy them.
12:36 And what is important is that the thought experiment
12:40 I want you to keep that expenditure is gonna be unchanged so this is not about a fiscal
12:44 multiplier where the government increases overall expenditure and then
12:47 there are positive economic outcomes is spending the same money
12:50 just changing the composition of firms from which I buy
12:53 and in principle you can think of this policy as being very cheap.
12:56 Governments don't have to spend more
12:59 at least directly.
13:00 They just have to make sure that they change
13:01 the composition of firms from which they buy.
13:03 So this is the thought experiment.
13:05 Of course one of the examples is promoting the participation of small firms.
13:09 Then the question is what is gonna be
13:11 the impact of this policy on aggregate outcomes?
13:14 So
13:15 what I'm gonna say is pretty obvious,
13:16 but if you want to look at the net effect of the policy first,
13:19 you have to identify the channels through which
13:21 the policy generates a positive effect,
13:23 and this should be aligned in principle with
13:26 the rationale from the policymaker point of view.
13:28 This should be consistent with what the policy
13:31 maker has in mind that this is gonna,
13:33 that this is gonna tackle.
13:35 Identify the channels through which the policy generates negative outcomes
13:39 and this may be more related to unintended consequences.
13:42 Maybe some of the policies that governments
13:44 policymakers are implementing are generating effects they were not expecting.
13:49 And of course you're gonna quantify the trade off and
13:51 if you find that the former is bigger than the latter
13:53 then the effect of the policy is positive and we can think of a policy that has been
13:58 uh
13:59 you know um
14:00 rightly implemented.
14:02 OK,
14:03 now I'm gonna try to give you a very general framework that
14:05 I think is gonna help us to think about that thought experiment.
14:09 Think of the government as a final producer
14:11 and
14:12 actually this is the way national accounts think about the government.
14:15 OK,
14:16 so governments produce health care,
14:18 they produce
14:19 education,
14:19 they produce
14:20 safety
14:21 along in that process they use intermediate input so public procurement is the
14:25 system governments have to acquire all
14:27 these intermediate inputs they need in order
14:30 to produce these these public goods.
14:32 Now
14:33 if you just think about the government in its role as a final producer
14:37 only that role
14:39 in principle what the government should do is to choose the set
14:41 of suppliers that maximize the output
14:44 given the constraint governments have a budget
14:46 if I just want to produce
14:48 as efficient as possible I'm gonna take that budget
14:51 and spend it optimally just by choosing suppliers
14:54 that can produce that at the lowest possible price
14:56 that would be in principle the main objective.
14:59 But in practice we all know governments and probably rightly so
15:02 they have additional objectives they only
15:05 they care not only about
15:06 profit maximization or cost minimization they care about other things
15:10 for example they may care about fixing an economic distortion.
15:13 For example,
15:14 financial constraints there may be
15:16 an economy in which firms are financially constrained,
15:18 so governments may want to internalize that
15:20 when deciding from which firms they want to buy.
15:24 Some governments don't
15:25 only care about the price index of consumers,
15:28 but they may care about creating jobs locally.
15:31 And again they may want to use the procurement system to somehow
15:36 uh tackle this objective which is
15:38 you know improving local economic activity.
15:41 So in principle
15:42 what I want you to keep in mind is that
15:44 it is possible and probably rightly so
15:46 that governments are gonna be willing to distort
15:50 their supplier choices so they can become more
15:52 inefficient from a cost minimizing point of view.
15:56 In exchange of
15:57 on tackling those additional objectives they may have in mind,
16:01 so that's kind of the framework I'm gonna be using throughout my presentation.
16:06 And what I'm gonna be repeating all the time is that this is
16:09 gonna imply a trade off there is gonna be a trade off between
16:13 increasing private sector performance by tackling
16:16 these potential distortions and additional objective
16:19 versus government efficiency so the government is gonna be willing
16:22 to be a less efficient final producer as long as this comes with
16:26 some benefit which may be coming from the performance on the private sector.
16:30 OK,
16:30 in the case of the Buy American,
16:31 as I mentioned before,
16:33 the government is already telling you how much
16:36 is willing to pay more
16:38 just
16:38 to create local jobs in that case was around 20%.
16:43 So this is just a graph
16:46 to repeat this framework with some equations and some nice arrows.
16:51 Government final producer
16:53 is gonna produce public goods and services which is YG
16:56 and the set of M's are gonna be suppliers,
16:59 OK,
17:00 potentially this set of suppliers,
17:02 uh,
17:03 is the set of all firms in the economy.
17:04 So in principle,
17:05 the government could actually buy from any firm in the economy in principle.
17:09 These are the suppliers which are just firms,
17:12 uh,
17:12 indexed by 12 up to G.
17:16 So what is the buy local policy?
17:17 The buy local policy is to change this
17:20 composition
17:21 and to make it more biased towards local firms.
17:24 Buy small is the same but making it more biased towards small firms
17:28 and what are gonna be the positive
17:30 effects of these policies?
17:31 Well,
17:32 these firms may actually do better in the private sector.
17:35 Why?
17:36 Because they have a new demand
17:38 from the government they may overcome some constraints.
17:41 So by changing the composition that's gonna affect the way
17:44 these firms behave in the private sector this may lead to positive effects.
17:49 But at the same time if you change the composition of firms from which you buy
17:52 and these firms are potentially
17:54 uh different in terms of productivity that's gonna affect the
17:57 prices they charge potentially affecting the price index of the government
18:00 and if we think that the budget of governments is fixed,
18:03 changes in the price index is gonna
18:06 inevitably affect how much real output the government is producing
18:10 so this is just the same trade as I was mentioning before.
18:14 OK,
18:14 so what I'm going to do next.
18:17 Is to give you two particular applications that I think fit really nicely into the
18:21 motivation in the sense that they speak to policies that uh I was emphasizing before
18:27 but also fall nicely into the framework that I just
18:30 described of course these two applications are only 2.
18:34 Uh,
18:34 and there are many other possible applications that I
18:37 still think would fall into this trade-off that I
18:40 talked to you before.
18:42 So the first application is gonna be
18:44 uh what are the aggregate effects
18:47 of buy small policies which is joint work with uh Julian Di Giovanni,
18:51 Pregina Enrique Mora,
18:52 and Giuseppe Johnmas.
18:54 And the second application is gonna be uh based on buy local.
18:58 It's gonna,
18:59 I'm gonna give you details later,
19:00 but it's gonna be
19:01 trying to understand to what extent even in a context in
19:04 which governments are not supposed to discriminate against non-local firms,
19:08 even in a context where there is
19:09 regulation going in the other direction,
19:12 there seems to be evidence that governments still
19:14 want to implement some sort of local bias
19:17 and I'm gonna try to tell you what would be the
19:19 cost of these local bias governments are implementing in Europe.
19:23 And this is joint work sorry with Marta Santamaria from from Warrior University.
19:28 So
19:29 what are the common ingredients
19:31 other than falling into this trade-off that I was emphasizing before?
19:34 The common ingredients of these two research pieces
19:37 are the following.
19:38 First,
19:39 all these research policies or these research questions are informed
19:42 by very granular data and procurement at the transaction level.
19:46 And this is something I want to emphasize
19:49 at the bank we have access to many data sets potentially at the transaction level.
19:53 What a transaction means in the context
19:54 of procurement in the context of procurement,
19:56 a transaction level data
19:58 means that we can observe the government agency buying
20:02 the product or service the government agencies is buying,
20:04 and the firm from which the government is buying.
20:07 So this is what I mean by transaction.
20:08 A transaction is a project procurement level that
20:11 has been awarded to a particular firm.
20:14 This is gonna give us a lot of information because
20:17 if we have that information we can merge it with firm levels more standard data sets
20:22 and we can see
20:23 how
20:24 uh firm's participation in procurement may actually help them
20:28 in in becoming better in the private sector which is
20:30 precisely the positive effects I was emphasizing so in other words
20:34 this granular data is gonna help us to discipline
20:37 the potential positive effects
20:39 that I was mentioning before.
20:42 Then
20:43 another common ingredient across these two pieces is micro evidence,
20:46 um,
20:47 using this micro data at the procurement level to say something about how
20:51 buyers perform,
20:52 what are the buyer's choices in this case government choices,
20:56 and how these choices affect the sellers which are the firms.
21:01 And critically because we want to be thinking about this trade
21:04 off and we want to be running counterfactuals at least conceptually is
21:07 gonna be very important to have some theory that is gonna
21:09 tell us how to go from the micro evidence that I was
21:13 that I've been emphasizing
21:14 to macro implications which is
21:16 ultimately the last objective we want to have
21:19 what is gonna be the effect of a particular policy in a country's GDP
21:22 in a government sector performance,
21:24 and so on and so forth.
21:26 So let me jump into the first application.
21:29 So the question we're gonna ask
21:31 in this particular application falls into the buy small umbrella,
21:34 but in particular the question is gonna be.
21:38 What are gonna be the aggregate effects of
21:40 a policy that actually targets small firms?
21:43 Very concretely.
21:44 What is the what is the political context?
21:46 What is what is what made us think about this potential question?
21:50 The context is the following so
21:52 all European countries,
21:53 all countries that belong to the European Union
21:56 have to adopt
21:58 European Union level regulation and procurement.
22:00 So the European Union comes up with directives
22:03 about the public procurement system and all countries have
22:06 limited time
22:08 to actually adopt those directives into the regulation.
22:12 So
22:13 even though there was no explicit
22:15 regulation
22:16 forcing agencies to discriminate in favor of small firms,
22:21 whether the governments or not should discriminate against uh sorry in favor of
22:24 a small firm was at the core of the debate for many years.
22:28 OK,
22:28 so for example there was a lot of encouragement
22:31 uh to public agencies to divide big lots into smaller lots
22:34 in a way that small firms would actually participate more often.
22:38 But there was nothing a reservation policy nor a big preferences
22:42 uh as the ones that I show you for the case of Indonesia or the US so the debate was
22:47 should we actually take one of those policies that are
22:49 prevalent around the world and implement them then in Europe
22:52 in a way that governments are more forced to buy
22:54 from small firms that was kind of the policy debate.
22:57 So the question we were trying to ask is precisely that what
22:59 would happen in a context in which there
23:01 is potentially positive effects and potentially negative effects.
23:06 So what we do is we start with a framework which is pretty standard.
23:09 It's a framework in which firms are heterogeneous in
23:12 productivity so you may think about an economy in which
23:14 there's firm dynamics firms get productivity shocks and they grow out of it,
23:18 um,
23:19 so they are heterogeneous in this productivity.
23:22 And they have access to a production function
23:24 which is very standard just for simplicity,
23:26 let's assume they use capital
23:28 and they mix capital with this productivity component which
23:31 is S and you can think of A as being
23:34 one of the sources of heterogeneity of these firms that is exogenous to firms,
23:37 firms just get shocks and get more or less productive.
23:41 From the demand perspective,
23:42 which is very standard is the fact that all firms sell
23:45 to the private sector at no cost just for simplicity.
23:47 So all firms because they produce differentiated varieties,
23:50 they're gonna have some market share in the private sector,
23:52 so they're gonna be,
23:54 uh,
23:54 producing and selling to the private sector.
23:58 What is more of an innovation
23:59 is the fact that firms have the ability to sell to the government
24:04 by paying a cost and in particular what we're gonna assume is that firms can invest in
24:08 increasing the probability of selling to the government
24:11 they can hire people
24:12 to make sure they comply with regulation.
24:15 They may prepare the bids correctly.
24:16 They may prepare for the auction,
24:18 but there's always some probability that they fail and they don't
24:20 get the contract they can affect that probability that's the idea.
24:25 What about finance?
24:25 So we're gonna assume that the distortion that is making the policy potentially um
24:32 good for the economy is the fact that firms are gonna be financially constrained.
24:35 So firms are gonna solve a dynamic problem in which they're gonna
24:39 borrow over time and they're they're gonna accumulate net worth.
24:42 Why do they want to accumulate net worth?
24:44 Because the level
24:45 of capital they can have an equilibrium is gonna be
24:48 constrained by the amount of net worth they have.
24:50 So this you can think of this as a collateral constraint.
24:52 OK,
24:53 so that's gonna provide incentives for firms to
24:54 accumulate net net worth in this economy.
24:57 Something we're gonna try to innovate as well based
24:59 on the evidence is that firms not only can
25:02 borrow against their network,
25:03 their tangible capital or their tangible assets,
25:06 they can also borrow against their
25:08 current revenue
25:09 and in particular we're gonna assume
25:11 that government contracts potentially but also
25:13 profits or revenue from the private sector can be used as a collateral
25:17 so there is some degree of pledgibility of these revenues.
25:21 So
25:22 what are the two
25:23 main advantages of selling to the government?
25:26 Why do firms want to sell to the government in this model?
25:28 First,
25:28 because there is another source of demand,
25:30 so they can make just more profits by selling to another demand.
25:33 They can use those profits to accumulate net worth which
25:36 is gonna allow them to borrow more in the future
25:39 because they're gonna have more collateral in the future.
25:41 But also because just on impact because government contracts
25:45 are pledgible they're gonna be able to borrow even more
25:48 even just the same year where they get
25:50 the contracts those are gonna be the two incentives
25:52 and the two channels through which selling to the
25:54 government is gonna be beneficial for the firms.
25:58 So
25:59 before telling you what is the role of the government in
26:01 this model or the way we think about the government,
26:03 let me tell you what is,
26:05 what is the impact of financial frictions,
26:06 what is the the kind of distortion the government is gonna try to handle.
26:11 So what is gonna happen is that firms that may be potentially very productive
26:15 but that have very low levels of net worth for example young firms,
26:18 you may be a very.
26:20 Productive firm,
26:20 but when you start you don't have enough level of assets
26:23 because the financial constraints you cannot borrow against those assets because
26:26 you don't have them and then you are inefficient this mode.
26:29 So what that's gonna imply is that
26:30 these firms are gonna be financially constrained.
26:32 There's gonna be a set of firms in
26:33 equilibrium that are gonna be financially constrained,
26:35 meaning that
26:36 their level of capital is too low.
26:39 And
26:40 their marginal productivity of capital is too high.
26:43 They have a lot of potential and if you were to
26:45 give them one unit of capital they would produce a lot,
26:47 but they
26:48 don't have capital because they're financially constrained.
26:52 Once you aggregate
26:53 these effects at the micro level,
26:55 something you're gonna find is that there is an aggregate.
26:58 Capital and TFP laws.
27:00 Why?
27:01 Because
27:01 you're gonna have a bunch of firms that are
27:04 smaller than optimally,
27:05 so the amount of capital in the economy
27:06 in equilibrium is gonna be inefficiently low,
27:09 and you're gonna have a bunch of firms that are financially constrained.
27:11 They have a lot of MRPK,
27:13 so TFP is low because you would like to relocate capital
27:16 from a lower uh marginal productivity,
27:18 uh,
27:19 firms to the high productivity firms.
27:21 And this is how somehow you're gonna get a GDP loss in the aggregate.
27:27 So what is the government gonna do?
27:28 The government is again is gonna be a final producer that
27:31 is gonna produce final goods that consumer is gonna consume.
27:35 And
27:36 along that process is gonna buy intermediate
27:38 inputs from these private sector firms.
27:41 So what we're gonna assume is that the
27:42 government has access to some policy instruments that allow
27:46 to affect procurement outcomes so somehow the government is gonna
27:49 have some policy tools that is gonna affect them to affect
27:52 the share of expenditure that is allocated to small firms versus large firms.
27:57 And how is done in equilibrium?
27:59 Well,
27:59 you may think about the policies I told you at the beginning or just simply a reduced
28:03 from way of making relatively more or less costly
28:05 for small firms to participate in procurement contracts,
28:08 OK?
28:09 And again the objective,
28:11 although we're not gonna be thinking about optimal policy but
28:13 the underlying objective of the government is gonna be.
28:16 To change the procurement system in a way that is gonna potentially decrease
28:20 the GDP loss that is actually coming from the presence of financial friction,
28:24 so in a world in which
28:26 financial frictions don't exist
28:28 then the role of cost minimizer
28:31 will be the only role the government has in this economy and then
28:33 what the government should do is buy from the most productive firms,
28:36 but because there are financial frictions there may be a way.
28:39 Of distorting the set of firms from which it's buying
28:42 in order to just attack these financial frictions that emerge from
28:45 from this distortion.
28:48 So then
28:49 same question but more concrete in the context of this model,
28:53 what would be the effect
28:54 of targeting the smaller firms more aggressively in this economy?
28:59 So
28:59 coming back to the policy context this would be
29:02 Spain.
29:04 In the year 2016 where there was no regulation,
29:07 no explicit regulation in favor of small firms trying to think about where
29:11 if Spain were to implement
29:13 what the European Commission has been debating
29:15 in the European parliament.
29:18 And the way we do it is by.
29:20 Decreasing the size premium.
29:21 What is the procurement size premium?
29:23 The the procurement size premium is that at a given point in time
29:26 you look at what is the average firm size
29:28 of firms that sell to the government versus not,
29:31 and what you can find is that firms that sell to the government are on average larger.
29:35 How much?
29:36 In Spanish data it's about 72%.
29:39 So what we do is to change this policy tool in a way.
29:42 That the composition of firms
29:45 from which the government buys is more biased towards the small firms
29:47 in a way that reduces this premium from 72% to 50%.
29:50 This is our counterfunction.
29:52 And of course this is very arbitrary,
29:55 but the point we want to make is that this seems
29:57 like a very cheap policy
29:59 is just changing the composition a little bit of the firms from which you buy.
30:04 Keeping expenditure unchanged.
30:06 Again,
30:06 same tradeoff as before,
30:07 how does the trade of operate in this model?
30:10 Well,
30:11 what is gonna be the positive effect on the private sector?
30:13 The positive effect is that.
30:16 When firms become
30:17 government suppliers,
30:19 they're gonna be able to overcome these financial constraints faster.
30:22 Why?
30:22 Through the two channels I mentioned before,
30:24 they're gonna make more profits because they,
30:26 they,
30:26 they have a new customer,
30:27 so they're gonna sell more,
30:29 so they accumulate those profits so they can actually,
30:32 uh,
30:32 have more capital in the future and borrow more in the
30:34 future but also directly because these government contracts have pledgibility.
30:40 So that's what it's gonna do is to make the self financing channel more powerful.
30:43 What is the self financing channel?
30:44 The self financing channel is a
30:45 channel emphasizing the macro development literature
30:48 where even in the context of financial constraints,
30:51 if firms who are productive today know they're gonna be productive tomorrow,
30:55 they're gonna accumulate assets over time.
30:57 So at the end
30:58 the aggregate effect of financial constraints is gonna be lower
31:01 than in a context in which firms cannot accumulate capital.
31:04 So what
31:05 the government is doing in this case is accelerating this self-financing channel,
31:08 making firms to overcome constraints much faster.
31:13 What,
31:13 what kind of evidence do we have in the data that actually tell us this may be going on?
31:17 How do we discipline this positive channel?
31:20 Here I'm just showing you what is the effect of becoming
31:23 a government supplier in the data and as you can
31:25 see is that credit goes up by around 6%.
31:28 And the effect seems to be persistent,
31:30 OK,
31:30 so firms seem to be able to borrow more
31:33 even after the duration of the contract.
31:36 What I think is very interesting and speaks more directly
31:39 to to one of the mechanisms I was emphasizing before
31:42 is the fact that.
31:44 100% of the effect that we find
31:47 is coming from a type of credit that we
31:50 call non-collateral credit.
31:51 Non-collateral credit means that in
31:54 in the credit registry of the Bank of Spain you can see what was
31:57 the collateral that was used for the firm in order to get that loan.
32:01 And we know whether the firm use some tangible collateral or not
32:04 so what we find is that it's not the case that on impact firms
32:08 use their capital
32:10 to borrow more after getting a procurement contract it just simply
32:13 seems to be the case that firms use those contracts.
32:16 As collateral to actually get more credit.
32:19 And what I think that is even more interesting
32:22 is that when you look at the tangible collateral type of credit
32:26 there seems to be no effect on impact.
32:28 It's only later on
32:30 when firms seem to have accumulated enough capital
32:33 to actually use that collateral credit to grow
32:36 so this is speaking directly to the self financing channel
32:39 on impact firms get a contract
32:41 they seem to borrow more from these noncollateral credit
32:44 and at the same time they accumulate capital that allows them
32:47 to use that capital as as as collateral in the future.
32:51 So this is the type of discipline that is gonna
32:53 help us identifying the strength
32:56 of the self-financing channel.
32:57 There is much more in the paper.
32:58 This is just a summary.
33:01 So what about total sales?
33:02 Total sales seem to increase on impact.
33:04 This seems obvious.
33:05 Firms get a contract,
33:06 so total sales should increase.
33:08 Two interesting things.
33:10 First,
33:11 the effect seems to be persistent.
33:12 So even when the contract is gone,
33:14 there seems to be a positive effect of around 2% in total sales.
33:18 What I think is more interesting is that when you look only at private sector sales,
33:23 which is kind of the object of interest for in the
33:25 trade of like what is the performance in the private sector.
33:28 We see there seems to be a crowding in of around 2-3%
33:32 after 4 years
33:33 despite the fact that on impact there seems to be a crowding out.
33:37 So what is the reason for this crowding out which by the way has been
33:39 found in in in other context by some of our colleagues at the World Bank.
33:44 The reason why there is a crowding out in the data and it's also generated
33:47 by the model is because selling to the government allows you to borrow more,
33:51 increases your borrowing capacity,
33:53 but it doesn't solve the constraint completely.
33:55 So you can think of this as in the short run firms are financially constrained,
33:59 so if they want to sell to the government they have to
34:01 decrease a little bit how much they sell to the private sector.
34:04 But precisely because of this self-financing channel,
34:07 firms have the ability to grow and accumulate capital
34:09 that is gonna allow them
34:11 to actually have a crowding in
34:13 after 4 years.
34:16 So there is a negative effect that we find from the
34:19 model that is hard to link to any data evidence,
34:21 but I'm just telling you because it's operating in the model,
34:23 which is the fact that big firms,
34:25 those that seem to be negatively affected by the policy,
34:28 those that are now less likely
34:30 to obtain contracts
34:31 seem to accumulate less capital in equilibrium.
34:34 What is the intuition?
34:35 The intuition is that
34:36 one of the reasons why firms accumulate capital in this economy.
34:40 Is because they want to make sure that in case they get a demand
34:43 shock they are not financially constrained so
34:44 they do it the kind of precautionary savings
34:47 type of mechanism
34:48 in a context in which big firms know that the likely
34:51 likelihood of obtaining a contract is lower they just want
34:54 to accumulate less precautionary savings and that's why in equilibrium
34:57 there is less amount of capital.
34:59 This is something that is kind of an unintended consequence.
35:02 Uh,
35:03 and intended also from the point of view of the researcher because
35:06 this is not something we were expecting was not obvious to us,
35:08 but that seems to be operating in the model.
35:11 So the overall effect in the private sector it
35:13 seems to be positive and we think it's huge
35:16 again just by changing the composition of firms from which governments buy
35:20 by a little bit
35:21 you can create you can get an increase in in
35:23 in real output of the private sector of around 1.2%.
35:28 We find that to be very high.
35:30 That's the private sector that's one part of trade off.
35:32 What about the other part of the trade off?
35:36 Um,
35:36 as I was anticipating in my narrative,
35:38 there's gonna be a cost that the government is gonna have to pay
35:42 by changing the composition of firms,
35:44 and the idea is that by targeting small firms
35:47 and then we can think about what would
35:48 be alternative ways of implementing this policy,
35:50 but just given the way we seem to see that governments do it in the reality.
35:55 By
35:56 pulling from the set of firms that are small firms or sorry,
35:59 governments are actually end up buying
36:01 from firms that are less productive on average.
36:04 They may buy from firms that are small because they're very
36:07 constrained but also from small firms that are small because they're unproductive
36:10 so at the end of the day what it's
36:11 gonna happen is that governments are gonna end up paying
36:13 higher unit prices
36:15 for their purchases by buying from more small firms.
36:19 So that means that the price is gonna go up,
36:20 the price governments have to pay to produce,
36:23 um,
36:24 final outputs
36:25 and given that the budget is fixed,
36:27 what that's gonna imply is that the real provision of public
36:30 goods by the government is gonna go down by around 8%.
36:34 So once you take into account how how big the government is.
36:38 Versus the private sector and these two effects that I was mentioning
36:41 you're gonna get that on average the net effect on real GDP is around 0%.
36:47 OK,
36:47 so then
36:48 it's up to the government.
36:50 To somehow tell you whether I'm more interested in solving the private sector
36:55 distortion versus minimize or decreasing the provision
36:58 of public goods in the economy,
36:59 so there seems to be this trade of operating.
37:02 Which is very similar to the one I mentioned to you conceptually a few slides before.
37:06 Now what are the takeaways from this from this research
37:10 application?
37:12 There is a trade-off.
37:14 I've been repeating myself
37:15 a lot.
37:18 OK,
37:18 but how this tradeoff may depend on the level of development,
37:21 for example,
37:21 which I think is very important.
37:24 The stronger
37:25 increasing YP is gonna happen in two possible situations.
37:29 Situation number 1,
37:31 how big the distortion is.
37:33 If the distortion is huge,
37:35 then
37:36 the government has more incentives to actually participate
37:39 and fix the distortion
37:40 so in an economy in which there is a low CA,
37:43 which is
37:44 the extent to which firms can borrow against their assets
37:46 and the extent to which firms can borrow against their
37:48 private revenue,
37:49 if those parameters are lower,
37:51 it means that firms are gonna be more constrained in that economy.
37:53 It means that the government may have more incentives to actually.
37:56 Implement this policy so the less developed
37:59 the economies in terms of financial development,
38:01 the more likely it is that the exchange in YP
38:04 is higher.
38:06 Also it's gonna depend on
38:07 on the the the how powerful the tool is
38:11 if we believe in this case that
38:13 the government has more credibility or
38:16 higher better reputation that private sector agents
38:19 and hence is gonna
38:21 give pledibility to firms
38:23 that's gonna make the policy more powerful in the sense that that
38:26 selling to the government just by the reputation of the government is gonna
38:29 make banks to be more willing to.
38:32 Lend to firms that sell to the government
38:34 and has this self financing channel is gonna be even stronger.
38:38 Something that I think is super important
38:40 I just realized by
38:42 building all these slides that something I had not thought about
38:46 a lot but I think this is a data
38:47 moment that is crucial to understand all these policies,
38:50 which is what is the productivity gap
38:52 first I'm gonna tell you generally and then in the context of this
38:55 in this
38:56 this application
38:58 what is the productivity gaps between the firms that you are targeting.
39:02 Local small.
39:05 You name it versus
39:06 the productivity of the incumbents
39:09 that are selling to the government already.
39:12 How this came out
39:13 in our in our uh research piece it came
39:17 from calibrating the productivity distribution which tells you how different
39:21 are small versus large firms in terms of productivity.
39:25 But we could have more more direct evidence on how
39:28 more productive are the income of the actual suppliers to the
39:30 government versus those that will be benefiting from the policy,
39:34 and that's gonna tell you
39:35 to what extent the negative effect is gonna be big or small.
39:39 OK,
39:39 so that's kind of the takeaways from from this particular application.
39:44 I'm gonna tell you about the 2nd application,
39:47 then some final remarks,
39:48 and then I will conclude,
39:50 so in the last 10 minutes.
39:52 So
39:54 As I told you,
39:55 public procurement is in principle integrated in the EU
39:58 single market spirit.
40:00 Governments cannot discriminate against
40:02 non-local firms.
40:04 However,
40:05 when you look at government purchases in the EU,
40:07 there's highly,
40:08 highly geographically,
40:10 uh,
40:10 concentrated,
40:11 and there was,
40:12 there is very nice research piece by one of our colleagues,
40:14 Alan
40:15 here at the World Bank
40:16 showing that.
40:18 So how big these numbers are?
40:19 Well,
40:20 if you look at what is the share of procurement value that is awarded to
40:24 firms
40:25 within the same countries,
40:26 pretty much everything,
40:27 98%.
40:29 When you do that at the regional level,
40:31 what fraction of
40:32 procurement values awarded to firms or establishments that
40:35 operate within the same region is around 50%,
40:38 so it's a lot,
40:38 and this is for France and Spain,
40:39 but pretty much prevalent across all countries in the.
40:43 Actually,
40:44 despite the fact that in principle integration is total,
40:47 there seems to be a lot of effort in
40:49 in fixing some problem so
40:50 so regulators policymakers recognize that despite
40:53 the regulation being the right one
40:55 there seems to be something that is not working well.
40:58 So the question is what are,
41:00 you know,
41:00 are government agencies home buyers in the EU despite the fact that regulation
41:04 shouldn't allow them to do it?
41:07 And if so,
41:07 what is the effect of this bias in
41:09 explaining this lack of market integration in Europe
41:12 in procurement and what is the cost on government's
41:14 provision of public goods again the same object as before
41:17 how much is home buyers costing governments?
41:20 So there's gonna be one challenge.
41:23 Which is that
41:25 We're gonna try to infer what is the government's home bias,
41:28 which is kind of a preference parameter if you want.
41:31 And
41:33 Main challenge is the fact that in principle just the fact that markets are
41:37 geographically
41:39 not integrated is not telling you that this should be a home buyers.
41:42 One explanation we all know,
41:43 uh.
41:45 From working with trade malls is that there may
41:47 be bilateral frictions about which governments cannot do anything,
41:49 so governments are constrained in the sense that
41:52 there seems to be some transportation costs,
41:54 information frictions,
41:55 and governments
41:56 buy as much as they can from local firms
41:58 or from non-local firms because there are some technological
42:01 restrictions that
42:03 don't allow them to do it
42:04 and the other possibility is home buyers.
42:06 So how do we disentangle the two?
42:08 That's what the paper is gonna be about
42:10 and actually the policy implications are very different.
42:13 If we believe that the problem of the lack
42:15 of market integration in Europe is number one.
42:18 Then the EU is doing the right thing,
42:19 which is increasing information flows,
42:22 um,
42:23 making sure that there is good infrastructure,
42:25 making sure things are advertised,
42:26 and so on and so forth.
42:27 If we think it's about
42:29 2,
42:30 then it should be about making sure governments comply with the regulation more,
42:33 more,
42:34 more strongly.
42:36 So two novel strategies to identify home buyers.
42:39 I'm gonna be super brief on it and then we're gonna use the structural model
42:42 to tell you,
42:43 given that estimated home buyers,
42:45 what is the effect for market integration and what
42:46 is the effect on the provision of public goods.
42:49 And this something that is important,
42:51 this analysis
42:52 in this research piece is partial in the sense that we're gonna be abstracting from.
42:57 The positive part of the trade off.
42:59 So whatever number I found about the loss
43:01 in government efficiency coming from home buyers,
43:03 you have to trade it off
43:05 against
43:06 what would be the positive benefit at the local level of implementing bias.
43:11 So that's gonna be in that sense incomplete,
43:12 but
43:13 it will come in future research.
43:15 OK,
43:16 2 hypotheses,
43:17 um,
43:19 2 strategies to,
43:20 to identify home buyers very briefly.
43:23 The first one relies on the hypothesis that home.
43:27 May have a different meaning depending on the type of government you are
43:30 if you're a local agency then home means
43:34 municipality.
43:34 If you're a regional agency,
43:36 then home means the region.
43:38 If you're a national agency then home means the country.
43:42 So
43:43 in principle the hypothesis is that subnational governments
43:45 are gonna discriminate against firms from other regions
43:49 but national governments shouldn't care about the specific region from
43:53 you know
43:53 uh where the firm is from as long as the firm is from from the country.
43:58 So this is just to give you an idea
44:00 on how different types of governments behave differently.
44:03 So this is local expenditure shares
44:05 at the region level.
44:07 By government agencies based in each of
44:09 these regions but those that are subnational
44:11 so these are local governments,
44:13 subnational governments,
44:14 regional governments and those are national agencies
44:17 take the case of Barcelona in Barcelona.
44:19 There are 3 or 4 different types of police.
44:21 You have the local police,
44:22 the original police,
44:23 and the national police,
44:24 and all of them coexist.
44:25 They all of them have headquarters close by.
44:27 All of them by printing paper,
44:29 and in principle all of them have the same market access around around Spain.
44:34 But what you can see is that
44:35 subnational government systematically buy much more locally
44:38 than national governments do.
44:40 Of course this may be driven by sectorial composition like the
44:43 fact that subnational governments buy different goods and products than.
44:47 National agencies this doesn't seem to be the case.
44:49 In fact,
44:49 what we find is that
44:51 within each,
44:53 and this is only at the two digits,
44:54 but even if you do it super narrowly
44:56 within each very narrowly defined product,
44:58 what you find is a subnational government systematically buy
45:01 much more locally.
45:04 Again this is not solving the challenge I told you
45:06 before because they may be the case that uh um
45:10 you know distortions or frictions are somehow different
45:13 across these government types so we do our best
45:16 using the following strategy.
45:19 Imagine you have just this data which is
45:21 you have the the regional police in Barcelona
45:24 buying printing paper.
45:26 And what we see in our data is that this
45:29 agency buys much more from Plan B,
45:31 which is the local plan,
45:32 vis a vis Plan A,
45:33 which is the non-local plan.
45:34 Again,
45:35 if you were to have only this data,
45:36 you cannot tell whether this is home buyers.
45:38 It could be the case that,
45:39 uh,
45:39 the regional police agencies constrained.
45:42 It's just much more costly to buy from Plan A.
45:44 So what we
45:45 do is to compare the same
45:48 printing paper and same two plants,
45:50 so the same plant in Madrid and the same plant in Barcelona,
45:53 and we just see how the national police behaves relative to the regional police.
45:57 It's kind of a definitive is how much this agency buys locally versus non-locally
46:01 vis a vis another agency that may have
46:03 different incentives from the same two plants.
46:06 So if you believe that this
46:07 nat natural frictions like transportation cost
46:10 information,
46:11 which is a challenge we try to address,
46:12 and many others are constant within an origin destination
46:15 but
46:16 uh sorry,
46:16 are constant within
46:18 an origin destination across government agencies,
46:21 then you can identify a relative home buyers
46:23 between these two types of government agencies.
46:26 So the second strategy relies on a natural experiment for France.
46:31 So in 2016 the number of regions in France decreased from 21 to 2013,
46:35 so this is just the map.
46:37 So take the case of
46:38 high Normandy and Low Normandy.
46:40 So before 2016
46:42 high Normandy and Low Normandy.
46:45 Were different regions that would be the gray the gray border
46:48 after 2016 they became the same region
46:51 so what we can see is whether firms based or establishing based in low Normandy
46:56 started selling more to government agencies based on
46:59 High Normandy after the reform,
47:02 whether firms that became artificially local just by the changes,
47:05 whether they became more active in the marketing
47:07 in High Normandy.
47:09 And what is key is that these two strategies control for origin destination factors
47:14 that may capture these technological constraints governments may have
47:18 and then as a result.
47:20 We identify a wedge that we interpret as a home bias.
47:25 What are the results?
47:26 Strategy number 1,
47:27 strategy number 22 things in common.
47:29 There seems to be much more market integration
47:32 in this case
47:33 there seems to be much more market integration when looking
47:36 at national agencies' purchases than when looking at subnational,
47:40 uh,
47:40 purchases.
47:41 In the case of the French reform,
47:43 there seems to be
47:44 much more market integration among the regions
47:46 that became integrated after the reform.
47:49 In the case of France,
47:51 most of the action,
47:52 actually all of it was through the extensive margin is like the,
47:55 the,
47:55 the fraction of firms in low Normandy
47:57 that now started selling to government agencies based
48:00 in up Normandy increased by a lot,
48:02 but how much by around
48:03 actually a factor of 4.
48:06 And what is interesting is that actually even though
48:08 here we are not relying on comparing governments.
48:12 There seems to be the case that subnational
48:14 governments react much more than national governments do.
48:17 Which is kind of consistent with the other strategy
48:20 in which we were taking that hypothesis as given.
48:23 So this is just
48:24 this event study for the case of France so this is the.
48:28 The graph you should be looking at,
48:29 so this is the,
48:30 the effect on the fraction of firms
48:33 from a non-integrated
48:35 pre-region
48:36 to a
48:37 post integrated region
48:39 in the,
48:40 the share of firms that sell into that region.
48:42 So it's like kind of the extensive margin.
48:43 What is the share of firms in low Normandy that sell into abnormandy before and after
48:48 and what we see.
48:50 Is that despite the fact there was no free trade,
48:52 there seems to be a big effect afterwards that is differential across
48:55 government agency which again reinforces the
48:57 hypothesis for the strategy number one.
49:01 So through the lens of the model we take a trade model
49:04 that maps
49:05 this
49:06 home biases into changes in in in trade shares local expenditure shares
49:12 and that also is linked to the price index of the government and what
49:15 we find is that even though we don't have direct information on prices,
49:18 that's one of the limitations I will emphasize later
49:20 through the lens of the model this expenditure share.
49:23 Inform about what is the cost in the price index
49:26 and we get that the home buyers that we estimate
49:28 is costing governments around 8% of their production,
49:32 so governments coming back to the trade off before seems to be an 8% less productive
49:36 as a function of these home buyers.
49:39 Does it make sense or not?
49:40 Well,
49:40 this is a wedge,
49:42 so,
49:42 so one of the issues we had is that we're,
49:44 we're identifying a wedge
49:46 and we don't have data counterpart to just confront it
49:49 against.
49:50 So what we did was actually learning we didn't know about it,
49:53 but in the US there's a lot of local
49:54 by local policies imposed by state governments very explicit.
49:59 And what we find is for example in the case of South Carolina we find an 8%
50:04 uh preference for local firms and this maps very nicely into our result because the
50:08 South Carolina state government is explicitly say
50:11 I'm gonna be willing to pay 8% more
50:14 as long as the establishment is from South Carolina.
50:17 Of course it's not exactly the same object,
50:19 of course it's not the same context,
50:20 but this is saying that we're finding
50:22 in terms of magnitude something that is not crazy.
50:26 So main takeaways
50:29 Governments seem to be more locally biased than as reflected by actual policies,
50:33 so this means that this may be a lower bound
50:35 in the cost of the existence of home buyers if we
50:37 were to do this in a world in which there
50:39 is on top of this underlying home buyers more regulation,
50:42 maybe the effects.
50:44 Even bigger.
50:46 So reduced from evidence plus structural models significant
50:49 cost in terms of uh provision of public goods.
50:53 And again here the key is what is the productivity gap
50:56 between the
50:57 firms you are targeting in this case local firms
50:59 versus the firms that are already selling you,
51:01 which may be potentially the foreign firms
51:04 if foreign firms and local firms are similar,
51:05 then it doesn't matter.
51:07 The reason why I'm emphasizing this data moment a
51:09 lot is because I think it's relatively easy to measure
51:12 in the data with procurement data it shouldn't be that difficult.
51:16 If you have unit prices,
51:17 you may know in equilibrium what is the price paid by.
51:20 Governments of different types of firms come up with an estimate on how different
51:24 these prices would be if you were to target firms that are
51:27 observationally equivalent to the ones that you have already in your sample.
51:31 Few questions that are open
51:33 why are governments locally biased?
51:36 Potentially because of many reasons.
51:38 So in some work
51:40 uh with Maria Arnal we actually were trying to implement uh a pilot survey
51:45 to procurement officers in Spain just asking them
51:48 first whether they perceive
51:50 their purchases to be as locally biased as we can
51:52 see in the data so something nice is we can check
51:54 what they say with our administrative data and see
51:57 whether at least they're aware of what they're doing
51:59 and just trying to ask them what are the the
52:01 the reasons why they want to emphasize local economy maybe.
52:04 Uh,
52:05 they think this is the proper way to do,
52:07 uh,
52:07 because it's greener just to see what do they have in mind.
52:11 So far we failed badly,
52:14 so we send out this uh salary to around 4000
52:18 procurement officers and and we had
52:21 a response by 30 or something so like response rate
52:24 was pretty low so but but that I think this is
52:27 I think connecting this type of just just to see
52:30 how policy makers think in in in procurement is it would be important.
52:33 And then something that seems to
52:35 imply from from these findings.
52:38 Is that governments should buy more from
52:41 from non-local firms governments should import more
52:44 buy American makes no sense.
52:45 Government should go and buy from other countries.
52:49 But this may not be as obvious,
52:51 so something we're exploring with uh Debay and Bob
52:55 is whether when we actually see governments buying stuff,
52:59 whether we see that they
53:01 pay more or less relative to private sector firms.
53:03 It may be the case that as as as importers governments are not very efficient.
53:08 For some reason,
53:09 because they don't invest enough in finance suppliers because they
53:12 just get a long term relationships that are sticky for some reason,
53:16 even if they were to import from abroad,
53:17 it is not clear they will have this productivity
53:19 gains that I'm inferring through the lens of my model
53:22 because I'm missing some technological restriction that I don't have in my model.
53:26 So this is what I'm,
53:27 what we're trying to,
53:27 to implement with the back end model.
53:30 So
53:32 I'm a bit late coming back to the trade-off,
53:35 uh,
53:35 relatively extensive evidence on the former,
53:38 so I think we have a lot of evidence trying to see how firms react
53:41 after the shock.
53:42 We have very little evidence on the direct effect on prices,
53:44 so everything I told you
53:46 was relying on the structure of models and some micro data,
53:48 but it would be great to have this more direct evidence on prices.
53:52 And just for the sake of time I abstracted
53:54 from topics that I think are super important.
53:57 Um
53:59 And I think we'll be coming in the future
54:00 by many researchers within and outside the bank.
54:03 um,
54:04 thanks a lot for your attention.
54:19 Thanks Manuel for that really interesting
54:22 and,
54:22 and very clear presentation.
54:24 Uh,
54:24 I feel like we all learned a lot.
54:26 I,
54:26 I,
54:26 I certainly did.
54:28 Um,
54:29 Uh,
54:30 uh,
54:30 for now,
54:31 let's turn it over to our discussant.
54:32 Um,
54:33 Majed,
54:34 um,
54:35 can we put the
54:38 Meeting on the screen.
54:45 Well,
54:45 thank,
54:45 thank you very much,
54:46 uh,
54:46 Dion,
54:47 for inviting me for,
54:48 um,
54:49 this,
54:50 uh,
54:51 very nice,
54:52 uh,
54:53 topic to talk about it and also for a good presentation by uh
54:59 Manuel.
55:00 Um,
55:01 as,
55:02 as we can all see,
55:04 procurement
55:05 is really becoming a,
55:07 a strategic function in the recent,
55:10 uh,
55:10 recent years.
55:12 And mainly the driver for that is the huge volume
55:16 going through public procurement.
55:20 It is estimated about $13 trillion globally
55:25 goes through public procurement.
55:28 And that's what really is attracting many researchers
55:34 in the recent years to really tackle the uh procurement aspects.
55:41 Now in the past it was
55:43 the procurement was more tactical function.
55:47 But with the introduction of the electronic
55:50 public procurement where data can electronically be generated.
55:55 And with many countries
55:58 that have the EGP.
56:00 The data became available to researchers,
56:03 and they start to really see the value
56:06 and the beauty of
56:08 analyzing procurement.
56:11 And let me use a term
56:15 I personally generate,
56:17 which is the procure
56:19 procurometer.
56:21 Um,
56:22 and this is really,
56:23 uh,
56:24 a measure
56:25 of how,
56:26 in my,
56:27 in my view,
56:28 how procurement is becoming more and more strategic function.
56:33 So the more we see.
56:36 People like
56:37 Manuel economists
56:39 getting into the procurement data and they start
56:42 to analyze the data and publish paper,
56:45 the more we understand the procurement is becoming a strategic function.
56:50 And I can tell you that we are really today in very great position.
56:57 In the bank and globally,
57:00 where senior management
57:02 in the bank and uh and outside the bank,
57:05 they talk more and more.
57:08 About the procurement
57:10 because
57:11 it is really obvious
57:13 being a tool to address country policies
57:17 and the global goods
57:19 in different aspects.
57:20 And here,
57:21 Manuel,
57:22 he talked about two aspects
57:24 buying a small
57:26 and buying local.
57:28 But there are many areas that researchers
57:32 and the procurement specialists
57:35 talking about these days,
57:37 including
57:38 the World Bank,
57:39 where we stress on
57:41 different aspects
57:42 like the sustainable
57:44 public procurement
57:46 and the domestic preference.
57:50 Now with,
57:51 with this,
57:51 um
57:53 when you talk about uh these aspects related to policies,
57:58 we need to be
58:00 very careful
58:02 when we use the data.
58:04 The data in electronic form is good and available,
58:09 but if it is not really vetted and uh uh not reliable,
58:15 it would lead
58:17 to
58:18 the opposite
58:19 results.
58:20 And I can tell you
58:22 with our
58:23 economist,
58:24 when I was with the Global unit in,
58:27 in prosperity.
58:29 We studied the impact of a procurement.
58:33 On cross-country trades.
58:36 And we
58:37 use the economists,
58:39 they use
58:40 government data
58:41 that available on the websites.
58:45 And unfortunately,
58:47 the results came in the opposite direction,
58:50 where it shows,
58:51 for example,
58:52 that if you join
58:54 the uh the uh
58:57 Government procurement agreement,
59:00 then the cross-country trade will be going
59:04 in the opposite direction,
59:05 which is not expected.
59:07 So we knew that the data set we are using,
59:10 it's not accurate.
59:12 Unfortunately,
59:13 many economists,
59:14 they use these data,
59:16 and there are many papers based on this data
59:19 in the literature.
59:21 So it's very important that
59:24 we vet the data.
59:26 Now coming to the two subjects
59:29 that Manuel tackled,
59:31 which are really
59:33 trending these days in many countries,
59:37 and
59:38 mainly because
59:40 Super,
59:42 super countries,
59:43 developed countries,
59:45 they are really using and
59:47 they bluntly in public advocating on using them.
59:52 For example,
59:53 buying local,
59:54 hiring local,
59:54 buy Americans,
59:56 hire Americans.
59:57 You hear it everywhere.
59:59 You go to India,
1:00:01 they use
1:00:02 uh make in Country A
1:00:05 as a preference.
1:00:06 And,
1:00:07 um,
1:00:07 and many other countries,
1:00:09 they,
1:00:09 they use these,
1:00:11 these kind of things.
1:00:12 Now,
1:00:14 Using these aspects would have impact,
1:00:17 as Manuel
1:00:19 mentioned.
1:00:20 And there is a merit for government
1:00:24 to use any of those policies
1:00:27 to address certain needs in the country.
1:00:29 So for example,
1:00:31 when it comes to sustainable
1:00:33 public procurement,
1:00:34 the government could be concerned with
1:00:38 a certain
1:00:39 group of people within the country,
1:00:42 and they would like to give them a preference so that they can really help them.
1:00:48 In
1:00:49 these local communities to live
1:00:51 better life,
1:00:53 so they can give them a preference.
1:00:56 And um
1:00:57 you have small and medium enterprises,
1:01:00 women-led small and medium enterprises.
1:01:03 You may want to encourage
1:01:05 women to get into the market more and more and more.
1:01:09 So yes,
1:01:10 there are merits on it.
1:01:11 However,
1:01:13 And I,
1:01:14 and I heard it from an economist
1:01:17 that we need to look for value for money,
1:01:20 for policy we use it.
1:01:22 So when a country like Spain
1:01:25 um uh set 8%.
1:01:28 And they agreed to pay 8% more.
1:01:32 When someone would need to say why I used 8%,
1:01:35 why,
1:01:36 why I did not use the 20%,
1:01:39 and why I did not use 2%,
1:01:41 and I think it is very important to look at the aspects of value for money.
1:01:47 Now,
1:01:47 when we talk about value for money,
1:01:50 it's not enough to talk about
1:01:53 the money.
1:01:54 Because there are other aspects that benefit,
1:01:57 and the economists
1:01:59 with the specialized people working together,
1:02:04 they can really factor in
1:02:07 to find the beauty of setting these,
1:02:10 these policies.
1:02:12 And uh in some countries also,
1:02:15 they have these kinds of policies
1:02:18 targeting their local suppliers.
1:02:21 But
1:02:21 it may be
1:02:23 hurting the local suppliers.
1:02:25 They may discriminate among
1:02:27 the local suppliers.
1:02:28 They think they are doing it in the right direction,
1:02:31 but it has
1:02:31 a negative impact on the local,
1:02:34 local suppliers.
1:02:36 Now,
1:02:37 that is really an aspect,
1:02:39 but
1:02:40 at the,
1:02:40 the other hand,
1:02:41 we need to look at
1:02:43 Issues that impact the project implementation
1:02:47 when it comes to public procurement.
1:02:49 So for example,
1:02:51 we in the World Bank,
1:02:52 we see when you have a small,
1:02:54 small contracts.
1:02:56 And the government does not have really enough capacity,
1:03:00 you will see lots of delays in these contracts.
1:03:04 So one way
1:03:06 Management is talking about it.
1:03:08 OK,
1:03:08 how about we lump sum?
1:03:11 We lump these smaller contracts into bigger contracts
1:03:15 so that the efforts we are doing.
1:03:18 will be similar
1:03:19 and
1:03:20 we're going to push lots of money going,
1:03:22 going through.
1:03:24 So you could see some
1:03:25 contradicting
1:03:27 the practices
1:03:29 when we talk about government procurement.
1:03:32 So
1:03:33 I'm,
1:03:33 I'm citing this because it's important to look at the policies
1:03:37 and it's important to look at the analysis of the data,
1:03:41 and our economists can really help us
1:03:44 in,
1:03:44 in doing,
1:03:45 uh,
1:03:45 doing that.
1:03:48 Um,
1:03:49 One time I asked an economist,
1:03:51 the vice president,
1:03:53 what do you think procurement has to do with
1:03:58 With the firms,
1:04:00 with the private sector,
1:04:01 and he answered by one single word growth.
1:04:05 So when we talk about policies that address small
1:04:11 We
1:04:11 are looking for those small firms to grow and get into the bigger market.
1:04:17 And that is really something maybe the economists can help us
1:04:21 in,
1:04:22 in doing that.
1:04:24 What I wanted to conclude with is
1:04:28 We are very glad that we are tapping into
1:04:31 research area now in procurement where we see it as
1:04:35 a
1:04:36 strategic function.
1:04:38 However,
1:04:39 we still need lots of research and the studies in this area,
1:04:44 as Dion mentioned,
1:04:45 that this is a new area for the economists
1:04:48 to really research and study.
1:04:52 So I'm just throwing in some,
1:04:54 some areas that can enlight your thinking.
1:04:58 For example,
1:04:59 what is the,
1:05:00 what is the connection between procurement and the social contract?
1:05:05 What is,
1:05:06 what is the impact of delays in a
1:05:08 procurement and contract execution on the economic benefits?
1:05:12 And I can tell you,
1:05:13 we did the research,
1:05:15 and it was really challenging
1:05:17 to find data even in our bank ICRs
1:05:23 on the economic benefits lost
1:05:27 due to delays in the project implementation in the bank.
1:05:31 However,
1:05:32 with the modest data we had,
1:05:34 We came with
1:05:37 A an economical model.
1:05:39 It is not published yet,
1:05:41 and it came out in the water sector as well as in the transport sector that about 1%
1:05:49 Of the capital cost of a project.
1:05:53 Is
1:05:54 the economic benefits lost due to delays of every single
1:06:01 month.
1:06:02 So if you're talking about $100 million
1:06:05 delayed for 10 months,
1:06:07 that's 10% of the $100 you're talking about $10 million
1:06:12 lost economic benefits
1:06:14 for the government and the citizens of the country.
1:06:18 Now,
1:06:18 when we talk about this kind of data and this kind of results
1:06:23 with ministries like Minister of Finance or economist,
1:06:27 then they will take it seriously
1:06:30 uh when it comes to um the impact.
1:06:34 We talk about corruption,
1:06:35 and corruption
1:06:37 in
1:06:39 the procurement and infrastructure could go
1:06:42 as high as 50% in some cases.
1:06:46 But we do not have really
1:06:49 good research on what is the impact
1:06:52 of corruption
1:06:54 on the economy and the economic benefits
1:06:58 government are losing and the citizens.
1:07:03 One area
1:07:04 that
1:07:06 came out in the recent years and during COVID-19
1:07:10 is the supply
1:07:12 chain system.
1:07:13 And if you have an inefficient supply system
1:07:17 that is impacting on the benefits and the prices
1:07:21 and all what we are talking about
1:07:24 can really
1:07:25 have different trend when it comes on
1:07:29 the analysis.
1:07:32 The impact of
1:07:34 the capacity building.
1:07:36 And what are the benefits of the capacity building?
1:07:39 This is an area where we have no clue
1:07:43 about it to justify
1:07:45 for governments
1:07:46 to spend the money.
1:07:48 To
1:07:49 uh uh skill enhancement of their procurement specialists.
1:07:54 On public procurement.
1:07:56 They are spending money,
1:07:57 but we are not telling them what is the economic benefits from doing that.
1:08:01 And thus they are pushing back
1:08:04 some countries in borrowing
1:08:06 to do capacity building.
1:08:08 I'm just citing some
1:08:10 areas that the economists
1:08:13 that really help us in a procurement in shedding.
1:08:17 Some economic
1:08:18 data analysis and results that can really help us in pushing agendas
1:08:24 that needed in the,
1:08:26 the recent day.
1:08:27 So the more we have
1:08:29 a kind of this research,
1:08:31 the more we will be better off when it comes to,
1:08:34 to procurement
1:08:36 and really provide
1:08:37 a good advice for governments on certain policies in,
1:08:42 in their
1:08:43 own countries.
1:08:44 Thank you very much.
1:08:51 Thanks,
1:08:51 Maj.
1:08:52 That was,
1:08:52 that was really great to hear from you,
1:08:54 um,
1:08:55 both some reactions,
1:08:57 but then also some ideas for,
1:08:58 for,
1:08:59 for future research.
1:09:00 I think,
1:09:01 uh,
1:09:01 I am gonna ask you to maybe respond a little bit,
1:09:03 uh,
1:09:03 but just to remind people,
1:09:04 if you're online,
1:09:06 uh,
1:09:06 please raise your hand or just indicate in the chat
1:09:08 that you have a question if you'd like to intervene.
1:09:11 Um,
1:09:12 maybe before,
1:09:13 and if you are in the room and you'd like to ask a question,
1:09:15 I,
1:09:15 we would ask that you use the microphone so that the people online can,
1:09:18 can hear you.
1:09:19 Um,
1:09:20 so.
1:09:21 Uh,
1:09:22 let's,
1:09:22 let's add some questions and then maybe do another round and then,
1:09:25 and maybe,
1:09:26 um.
1:09:27 And maybe Majid,
1:09:28 I'll come back to you after we've had some
1:09:30 discussion here and maybe if you have any further thoughts.
1:09:34 So let's,
1:09:34 let's take a couple of questions from the room
1:09:36 and then I don't see anything online yet,
1:09:38 but um we'll go there.
1:09:39 So
1:09:40 Miriam,
1:09:40 yeah,
1:09:40 hi,
1:09:41 um,
1:09:42 thank you for a great presentation and a great discussion.
1:09:45 I'm Miriam Bruun from um DC FP.
1:09:48 So I have a question about uh using procurement for small firms you
1:09:52 mentioned that sort of governments try to target market failures and sort of,
1:09:56 you know,
1:09:56 use the procurement to
1:09:58 alleviate the market failure,
1:09:59 but in some sense that seems to be second best like
1:10:02 with the example of collateral like why not
1:10:05 save the money and
1:10:07 implement like movable collateral registries or sort
1:10:09 of directly address the financial constraints.
1:10:11 I was wondering what your thoughts are on that.
1:10:16 Um,
1:10:16 hi,
1:10:17 uh,
1:10:17 thank you for a great presentation.
1:10:18 I actually have the same question.
1:10:19 I mean,
1:10:21 yeah,
1:10:21 I mean,
1:10:22 your textbook thing is you don't
1:10:24 eliminate distortion by creating another distortion.
1:10:26 I mean,
1:10:27 you,
1:10:27 so it seems like a very inefficient way of doing it,
1:10:31 uh,
1:10:32 but I take it that you,
1:10:33 you,
1:10:34 what you're saying is.
1:10:35 I mean,
1:10:36 people say it's bad to do like small firm preferences.
1:10:40 Well,
1:10:40 it's not that bad in the end because there's still some benefits.
1:10:43 I don't know if I'm interpreting this right,
1:10:45 but otherwise I agree with the previous question.
1:10:50 I so I have a question about the some of the results from the
1:10:53 buy small part of the presentation.
1:10:56 So,
1:10:56 so the question is um
1:10:58 how different is um
1:11:00 the effect of buy,
1:11:01 selling,
1:11:02 winning a contract with the government to winning a contract with
1:11:05 another firm?
1:11:07 Um,
1:11:08 so is it,
1:11:08 is it just the,
1:11:09 the contract that's important,
1:11:11 or is the government special in a way,
1:11:13 um,
1:11:14 in a sense that the revenues from that contract
1:11:18 are more able to be borrowed against?
1:11:20 And sort of a,
1:11:22 a follow up is how does this change depending on the firm and the government.
1:11:26 So
1:11:27 is it the case that if you get a contract with a large,
1:11:29 reliable multinational firm,
1:11:32 that's similar to selling to the government?
1:11:35 And then also,
1:11:36 if the government is,
1:11:38 um,
1:11:38 I mean some governments uh pay their bills on time and others do not.
1:11:43 So if you're selling to a government that maybe is in arrears,
1:11:47 is that
1:11:48 similar to selling to a
1:11:51 less reliable company?
1:11:57 Thank you.
1:11:57 Uh,
1:11:57 my name is Albernoez.
1:11:58 I'm in deck B,
1:11:59 and thank you,
1:12:00 Manuel for this presentation.
1:12:01 I think it's very ambitious to which I'm from
1:12:02 micro to macro into procurement and it's very helpful,
1:12:05 but I have a lot of questions.
1:12:07 Uh,
1:12:07 the first one is,
1:12:08 I,
1:12:08 I think,
1:12:09 a kind of basic
1:12:11 how,
1:12:11 how to reconcile that some of the data is expenditure and
1:12:14 some is like tender prices because you have the differences,
1:12:17 uh,
1:12:18 quite often that what was
1:12:20 won in a contract is not what was actually expended or like paid,
1:12:23 no.
1:12:24 And I'm saying this because this might have some strategies
1:12:27 depending if you're a small firm or a large firm.
1:12:29 A large firm might have the ability to
1:12:32 look like less productive at the tender stage,
1:12:34 but then because of our negotiations
1:12:36 or something that happens during implementation
1:12:38 is able to raise expenditure on the price,
1:12:41 and I mean because
1:12:42 I thought that maybe the first paper was more using
1:12:44 tender data and the second was more on expenditure.
1:12:47 And that might be important at least for,
1:12:49 for comparative into
1:12:51 some of the,
1:12:51 of these aspects.
1:12:53 And the second part also is that.
1:12:55 Uh,
1:12:56 although it's interesting this the,
1:12:57 the,
1:12:58 the composition that you can
1:13:00 like switch,
1:13:01 we know from procurement that there is not that much competition,
1:13:03 there is not that many
1:13:05 firms that compete within one market.
1:13:08 Some tenders don't go into competition at all
1:13:11 and maybe some of these reserve markets also don't
1:13:14 they go directly to 1
1:13:15 to 1 person.
1:13:16 So how,
1:13:16 how in your model do you account by.
1:13:19 Maybe the inability to switch
1:13:21 between
1:13:22 firms easily.
1:13:24 Uh,
1:13:24 because I assume you,
1:13:25 you,
1:13:25 you have a policy against which within.
1:13:28 A high productive with a low productive,
1:13:30 but maybe because there's no competition you cannot governments
1:13:33 don't switch,
1:13:34 um,
1:13:35 and,
1:13:36 and,
1:13:36 and one on the on the financial constraints,
1:13:39 um.
1:13:41 Like also,
1:13:41 also the large firms,
1:13:43 they try to compete for the market,
1:13:45 right?
1:13:46 And,
1:13:46 and I would assume they will have like more markets and small firms
1:13:50 um in terms of number of markets or volume if you want.
1:13:53 So how this will affect like or or if you took into account this into into the papers.
1:13:59 Thank you.
1:14:02 Thanks for the great presentation.
1:14:04 Uh,
1:14:04 so I have a major question about like objective function of government.
1:14:08 So like from the macro perspective we can think government as like try to,
1:14:13 uh,
1:14:14 minimize the cost,
1:14:15 but if we think or talk to people in the like a local government,
1:14:18 maybe that's not they're like a local,
1:14:21 uh,
1:14:21 public procurement specialist their perspective their like objective,
1:14:25 right?
1:14:25 It's not in their KPI that try to push them to.
1:14:28 Minimize the cost in the sense
1:14:30 and a lot like it depends on the like what's the structure of the public finance
1:14:34 if like they got funding from the central government they yearning to run
1:14:37 all those budgets by the end of the physical years which is super common
1:14:42 so I'm like thinking how to from
1:14:44 micro level to aggregate to macro uh like a macro level that's objective function
1:14:49 still holds and how that will affect the overall results for the papers.
1:14:53 Yeah,
1:14:54 thank you.
1:15:00 Hi Manu,
1:15:01 uh,
1:15:01 so you were saying that you expect the tool to be more
1:15:05 powerful perhaps in developing countries because
1:15:07 financial constraints may be larger.
1:15:10 It's kind of related to what Steven asked so
1:15:12 you hear all the time that in developing countries
1:15:15 governments are very famous for being very bad payers,
1:15:17 no,
1:15:18 to their suppliers and
1:15:19 not even if they are in arrears they just take.
1:15:21 Very long time to pay their suppliers,
1:15:24 so in that sense it is likely that the banks are internalizing that,
1:15:28 uh,
1:15:29 so the pledgeibility of these contracts
1:15:31 of the government contracts is very likely
1:15:34 low in developing countries.
1:15:35 So is there evidence on the impact of this tool
1:15:39 in developing countries and or this pledgeability?
1:15:44 Thanks.
1:15:46 Poor Manuel.
1:15:48 That was a long list.
1:15:49 If,
1:15:49 if,
1:15:50 if,
1:15:50 if this time we'll open it up again,
1:15:52 um,
1:15:53 but,
1:15:53 um,
1:15:54 sorry,
1:15:54 a lot of questions.
1:15:55 I let that go on for a long time,
1:15:56 but,
1:15:57 but,
1:15:57 um,
1:15:58 I'll give you a shot,
1:15:59 and then if,
1:16:00 if there's something I thought you should answer,
1:16:02 I'm gonna try to select those that I think are kind of more big picture,
1:16:05 um,
1:16:06 and then I'm super happy to discuss for as long as needed afterwards.
1:16:11 So,
1:16:12 uh,
1:16:12 Majeed,
1:16:12 thank you so much for the for the discussion.
1:16:14 I couldn't agree more with,
1:16:15 um,
1:16:16 with
1:16:17 all your comments actually so I selected two of the
1:16:20 comments that you made that I think are particularly important.
1:16:24 One is uh project implementation.
1:16:27 And this is something that we've been abstracting completely in
1:16:30 our research in the sense that our models are.
1:16:33 Able to capture these policies in a reduced from way
1:16:37 but we are aware of all the issues that
1:16:39 that uh appear when it comes to implementation like somehow
1:16:44 governments are not only
1:16:46 uh
1:16:48 final producers of inputs but they actually may be
1:16:50 constrained in the way they can implement these policies.
1:16:54 And the same policy
1:16:56 written down in regulation in one country may have
1:16:58 very different effects than in other country depending on the
1:17:01 ability of governments to implement it correctly and that's
1:17:03 something that we really have to think more about.
1:17:06 There was the issue of renegotiations
1:17:08 uh we're aware that renegotiations are a very important issue in in procurement.
1:17:13 And so far we haven't been able to incorporate these
1:17:15 renegotiations in our models but I think it's something that
1:17:19 is clearly limiting uh the you know the credibility of our results
1:17:23 so you just wanted we wanted to start with something that.
1:17:26 Uh,
1:17:27 abstract from all the details at that level and somehow provides kind of the first,
1:17:31 uh,
1:17:32 information about the main trade-offs that may be in place.
1:17:35 I think a super important question is the one raised about the second best.
1:17:39 Disclaimer,
1:17:40 this is not by any means
1:17:42 a research process in which we come up with,
1:17:45 OK,
1:17:46 there are,
1:17:46 there are financial distortions in the economy we're gonna try to convince you
1:17:49 that procurement policies are the ones that are gonna fix the constraint in the
1:17:54 in the best way.
1:17:55 Our approach at the beginning was more kind of a positive approach.
1:17:58 These are the policies that are around
1:18:00 every single government seems to be doing so.
1:18:03 And when you read the regulation and the reasons,
1:18:05 the rationale for these policies,
1:18:07 financial frictions at the core of this of this rationale.
1:18:10 So we wanted to have a framework to help governments to say,
1:18:13 OK,
1:18:13 given that you seem to be doing this pretty prevalently across all countries,
1:18:18 what are the effects that you may be actually having both negative and positive?
1:18:21 So this was our approach rather than trying to say something normative
1:18:25 about this is the policy you should implement or not.
1:18:27 It's more kind of informing policymaker about the trade-offs.
1:18:31 Then the Claudia and Steven,
1:18:34 I think that's
1:18:35 key,
1:18:36 but somehow reflected in our trade-off so I was emphasizing that
1:18:40 the extent to which buy small policies may be
1:18:43 uh more important depend on two factors.
1:18:45 Factor number one is the,
1:18:46 the,
1:18:46 the extent of the initial distortions you want to tackle,
1:18:49 which is financial,
1:18:50 uh,
1:18:50 constraints.
1:18:52 So in that sense if you just
1:18:53 take that,
1:18:54 uh,
1:18:55 partially that would imply that countries with higher levels
1:18:58 of distortions may benefit more from these policies.
1:19:01 But then I also mention
1:19:03 how powerful the policies.
1:19:05 I I was relating
1:19:06 that how our policies to the legibility of the government contract
1:19:10 so it may
1:19:11 very likely be the case that in developing countries.
1:19:14 Distortions are higher,
1:19:16 more prevalent,
1:19:17 so in principle the policy will have a a
1:19:19 higher impact but because the policy parameter is weaker,
1:19:22 which is the eligibility,
1:19:23 at the end of the day
1:19:25 things may not be as we expect.
1:19:26 So there's this horse race between the powers
1:19:29 of the policy tool versus the underlying distortions you are trying to solve.
1:19:34 So that's,
1:19:34 I think somehow captured in our framework.
1:19:37 Um
1:19:40 What you ask about multinationals.
1:19:43 We don't think there is
1:19:45 conceptually any difference between from the point of view of the firm
1:19:49 in selling to Walmart versus selling to the government.
1:19:53 Um,
1:19:54 we just think that governments seem to actually,
1:19:57 um,
1:19:57 target small firms with a policy objective in mind and we don't see
1:20:03 how
1:20:03 private firms may be put into the picture to actually
1:20:07 do the same job government seem to be doing,
1:20:09 but in principle there is nothing that makes a private sector firm
1:20:12 selling to Walmart to get the same benefits from selling to the government
1:20:15 again it comes to this pledgeibility parameter.
1:20:18 That in some context actually may be stronger for Walmart and from the
1:20:22 government because the government has no credibility so it depends on the context.
1:20:26 Um
1:20:29 What about the micro very micro aspects of procurement?
1:20:32 We are aware that
1:20:34 thinking about the 2nd prize auction,
1:20:36 1st prize auction,
1:20:38 uh,
1:20:38 score,
1:20:39 uh,
1:20:39 functions is almost impossible in a macro model,
1:20:42 and this is precisely our point.
1:20:44 Our point is that the procurement literature very successfully so was
1:20:48 making a lot of progress in analyzing auction by auction.
1:20:51 But when you look at the policies,
1:20:53 most of the policies are written in a very general way
1:20:55 that are not auction specific.
1:20:57 There are some exceptions,
1:20:58 but most of them is by American.
1:21:00 Which means
1:21:01 there is this preference gap
1:21:02 so we were trying to abstract.
1:21:05 In the most aggressive way from the micro aspects of procurement contracts
1:21:09 but in exchange capturing kind of the trade of the policy makers seem to have in mind
1:21:14 and tackle those
1:21:16 so we are aware that we cannot capture all the types of possible options that
1:21:19 happen in reality because our macro models are
1:21:22 not powerful enough and it wouldn't be tractable
1:21:25 but we're aware of the limitations and I
1:21:26 think we have to think much more carefully about
1:21:28 how these details coming back to implementation which is kind of a particular.
1:21:33 Uh,
1:21:33 case of,
1:21:34 of,
1:21:34 of the of more kind of macro details may affect our results and
1:21:37 to what extent we have to be careful in interpreting what we find.
1:21:41 Um,
1:21:42 Sorry if I missed
1:21:44 any of the.
1:21:45 No,
1:21:45 thanks,
1:21:45 Manuel.
1:21:45 Um,
1:21:46 let's do a lightning round.
1:21:47 We have 2 online and then Javi,
1:21:49 and then I'll go back to Madrid and then,
1:21:52 uh,
1:21:52 to,
1:21:52 to Manuel to close off because we only have about 5 minutes left.
1:21:55 So,
1:21:56 so,
1:21:57 uh,
1:21:57 Anna Kosalito.
1:21:59 Yes,
1:21:59 um,
1:22:01 yes,
1:22:02 sure.
1:22:02 Hi Manuel,
1:22:03 thank you so much for,
1:22:04 for the paper.
1:22:05 I was surprised a little bit to find that the relocation
1:22:09 of public resources from large to small firms doesn't actually have.
1:22:14 Uh,
1:22:14 any effect on,
1:22:15 on GDP.
1:22:17 Uh,
1:22:17 can we say that this is just a lower bound,
1:22:21 uh,
1:22:23 of the effect because if the productivity gap between
1:22:26 small and large firms is actually
1:22:28 low,
1:22:29 and that is,
1:22:30 uh,
1:22:31 basically in a setting
1:22:32 where
1:22:33 small firms are financially constrained,
1:22:36 then
1:22:36 if you move from
1:22:37 this setting to a dynamic one where actually you allow companies to
1:22:42 Um,
1:22:42 make investment in productivity enhancing activities,
1:22:45 you would expect to observe that,
1:22:47 uh,
1:22:48 basically,
1:22:48 small companies invest more and that sort of,
1:22:51 uh,
1:22:52 increases productivity and,
1:22:53 and output.
1:22:55 Uh,
1:22:55 and I don't know basically sort of the differences of
1:22:58 the dynamic effect between uh developing and advanced economies,
1:23:02 but definitely,
1:23:03 uh,
1:23:03 rephrasing a little bit the question that Stephanie says,
1:23:07 I think that
1:23:08 conditional that the government is distorting what is actually,
1:23:12 uh,
1:23:12 sort of the
1:23:13 type of suppliers that will help you maximize the,
1:23:16 your gains,
1:23:17 not only from a static point of view,
1:23:18 but also,
1:23:19 uh,
1:23:19 dynamically.
1:23:21 Thanks,
1:23:22 Ana.
1:23:22 uh Diego,
1:23:24 uh,
1:23:24 and please keep it brief.
1:23:26 Thank you.
1:23:27 Thank you.
1:23:27 Yes,
1:23:27 just a comment,
1:23:28 this is a very nice framework for
1:23:31 understanding the effect on women-led businesses.
1:23:33 Uh,
1:23:34 first,
1:23:34 because on the left side of the trade-off,
1:23:37 gender norms perhaps are a constraint as
1:23:40 big as financial constraints or correlated,
1:23:42 but might not be easy to address it.
1:23:43 Directly
1:23:44 and secondly on the other side of
1:23:47 the trade-off,
1:23:48 I wonder if you can use productivity as a sort of index,
1:23:51 taking into account the other conditions that
1:23:52 are relevant for procurement as much notice
1:23:55 as punctuality,
1:23:57 responsibility,
1:23:58 and quality of products.
1:24:00 Thank you.
1:24:02 Thanks Xby
1:24:03 very quickly,
1:24:04 um,
1:24:04 uh,
1:24:05 very quick,
1:24:05 so,
1:24:06 um,
1:24:07 so you focused on the,
1:24:08 you know,
1:24:09 these policies reflecting changes in the probability that that
1:24:12 a single that a firm will be selected,
1:24:14 so like buying local,
1:24:15 you know,
1:24:16 smaller firms having high probability,
1:24:18 but,
1:24:18 but I wonder,
1:24:19 you know,
1:24:19 more broadly
1:24:21 how the how you structure the,
1:24:23 the,
1:24:23 the,
1:24:23 the procurement.
1:24:24 So,
1:24:25 so you mentioned actually,
1:24:26 you know,
1:24:26 reducing the,
1:24:27 the.
1:24:28 You know,
1:24:28 dividing the,
1:24:29 the,
1:24:30 uh,
1:24:30 you know,
1:24:31 what you wanna buy into,
1:24:32 into,
1:24:33 into different,
1:24:34 different,
1:24:35 uh,
1:24:35 uh,
1:24:36 contracts so that,
1:24:36 so that smaller firms can,
1:24:38 so basically it's about
1:24:39 also,
1:24:40 you know,
1:24:40 reducing the cost of,
1:24:41 of,
1:24:42 of firms to,
1:24:43 to apply,
1:24:43 right?
1:24:43 So you know,
1:24:44 eliminating red tape.
1:24:45 I mean what are the,
1:24:46 what are the constraints there,
1:24:47 um.
1:24:48 And also,
1:24:49 you know,
1:24:49 again,
1:24:50 you know,
1:24:51 controversial,
1:24:52 but tackling the,
1:24:53 the,
1:24:54 the,
1:24:54 um,
1:24:55 the liquidity constraints,
1:24:56 I mean it could be,
1:24:57 you know,
1:24:57 something like advances,
1:24:58 but,
1:24:58 but then,
1:24:59 you know,
1:24:59 of course,
1:25:00 you know,
1:25:00 if you provide too much of an advance,
1:25:01 then,
1:25:02 you know,
1:25:02 the guy,
1:25:02 you know,
1:25:02 the firm can,
1:25:03 can basically,
1:25:04 but,
1:25:05 but again thinking more broadly in how you
1:25:08 structure these procurements so that,
1:25:09 you know,
1:25:09 you,
1:25:10 you target the right,
1:25:11 the,
1:25:11 the right,
1:25:11 the right firms I guess.
1:25:14 Thanks,
1:25:14 Rabi.
1:25:15 Um,
1:25:15 Majid,
1:25:15 did,
1:25:16 did you want to come in with any last comments before we,
1:25:19 I give the floor back to,
1:25:20 to,
1:25:21 to Manuel to close out?
1:25:24 Thank you very much.
1:25:24 I think the,
1:25:25 the last comment,
1:25:26 maybe I can comment on it,
1:25:28 um,
1:25:29 because it's a procurement process.
1:25:31 But uh,
1:25:31 as,
1:25:31 as Manuel mentioned,
1:25:33 there are uh different ways of addressing uh this uh
1:25:37 uh small uh buying small,
1:25:40 uh,
1:25:41 and also buying local.
1:25:43 Um,
1:25:43 one way is to,
1:25:45 to,
1:25:45 to set aside a certain percentage of the budget
1:25:49 where you would say the competition will be among only
1:25:53 the small and medium enterprises,
1:25:55 for example,
1:25:57 or you would,
1:25:58 uh,
1:25:58 you would say small and medium,
1:26:01 medium enterprises led by women.
1:26:03 So,
1:26:04 Big firms cannot compete.
1:26:06 If you,
1:26:07 if the firm is not led by women,
1:26:09 cannot compete.
1:26:10 So this is one way.
1:26:11 The other way is you give a preference,
1:26:14 like you are willing to pay a little bit more.
1:26:18 Just to
1:26:19 uh address woman-led
1:26:22 uh SMEs.
1:26:23 So if they call,
1:26:24 uh,
1:26:24 if they offer you a higher bid,
1:26:27 you may
1:26:28 accept it being the,
1:26:29 the winner within a certain percentage.
1:26:32 Now,
1:26:34 This has a merit.
1:26:35 The government in its policy has a need
1:26:39 to address that.
1:26:41 If,
1:26:41 for example,
1:26:42 you have
1:26:43 90% of your firms are small and medium enterprises,
1:26:47 and you have 10%
1:26:49 are winning
1:26:51 90% of the contracts,
1:26:53 then basically you are,
1:26:55 there is a distortion in,
1:26:57 in the private sector and you need to work.
1:27:00 To address the small and medium enterprises for,
1:27:03 for growth.
1:27:04 The same when it comes to social aspect,
1:27:07 you may have certain
1:27:09 aspects.
1:27:10 I,
1:27:10 I'll tell you one example,
1:27:11 in war areas.
1:27:13 You have lots of people
1:27:16 die,
1:27:17 and then certain communities will be vulnerable
1:27:21 for economic growth.
1:27:22 And if you do not really
1:27:25 give them a preference in the procurement in these local communities,
1:27:29 then people can come from outside and take the contracts.
1:27:33 And I can tell you one good example here
1:27:35 is in Florida when there was a hurricane.
1:27:38 The president at that time limited
1:27:41 the competition
1:27:42 to only bidders within the state of Florida
1:27:46 to address a need in that state.
1:27:49 Thank you.
1:27:50 Thanks Majid we've reached time,
1:27:52 so I'm gonna give Manuel the chance to respond,
1:27:54 but you don't have to
1:27:55 if you don't want to.
1:27:57 No,
1:27:57 just,
1:27:57 I mean this is kind of a um general answer to the specific question by Ana just.
1:28:03 Here I was trying to show you one
1:28:07 particular application in which
1:28:09 it seems to be that uh one of the objectives uh
1:28:13 governments are trying to tackle by buying from small firms
1:28:16 is attack financial distortions this is not by any mean.
1:28:20 Trying to emphasize that this is the only channel
1:28:23 through which uh the results may be affected.
1:28:26 This is not the only channel through which we
1:28:27 think governance may help or not help the economy.
1:28:30 This idea about uh firms making uh optimal decisions on innovation and
1:28:37 and R&D as a result of becoming government suppliers we think is very important and,
1:28:41 and we're just abstracting from it in this particular,
1:28:44 uh,
1:28:44 research project but but but we think there
1:28:46 are many channels through which this may happen.
1:28:49 Um
1:28:51 I'll leave it there.
1:28:53 Thank you.
1:28:53 um.
1:28:56 So thank you all for joining us today.
1:28:58 Thank you,
1:28:58 Majid,
1:28:59 for,
1:28:59 for joining us from Djibouti,
1:29:00 um,
1:29:01 and,
1:29:02 uh,
1:29:02 Manuel for the presentation.
1:29:04 Um,
1:29:05 uh,
1:29:05 the event is recorded and the materials will be put online.
1:29:08 I also want to take the opportunity to encourage you to,
1:29:11 uh,
1:29:12 use our newest,
1:29:13 uh,
1:29:13 uh,
1:29:14 toy,
1:29:14 which is the Deck Decoded podcasts.
1:29:17 Uh,
1:29:17 of which,
1:29:18 uh,
1:29:19 one of Manuel's papers is gonna be,
1:29:20 uh,
1:29:21 uh,
1:29:22 uploaded very soon.
1:29:23 So you can use the QR code to,
1:29:25 if you're on the World Bank internet,
1:29:26 uh,
1:29:27 you can download these podcasts,
1:29:29 uh,
1:29:30 based on various papers including Manuel's.
1:29:32 So thank you.
1:29:36 Thank you very much.
1:29:37 All the best.
- add-style
- lp-body-content