1
00:00:02,890 --> 00:00:05,810
Welcome to the CAC Research podcast.

2
00:00:05,810 --> 00:00:13,030
The series of conversations about projects taking place at the Centre for the Study of African Economies, University of Oxford.

3
00:00:14,020 --> 00:00:18,190
My name is Stefan Darko and I'm the director here of the CRC and a professor of

4
00:00:18,190 --> 00:00:22,570
economic policy at the Department of Economics and the Blavatnik School of Government.

5
00:00:23,020 --> 00:00:31,800
Last week, the Economic and Social Research Council of the UK hosted their Celebrating Impact 2023 prize ceremony.

6
00:00:31,810 --> 00:00:37,760
And I'm delighted to be joined today by the winner of the award for Outstanding Public Policy Impact.

7
00:00:37,780 --> 00:00:41,380
Kate Walk. Kate is a member of the CSA.

8
00:00:41,560 --> 00:00:46,090
She's an associate professor in Economics and Public Policy at the Blavatnik School of

9
00:00:46,090 --> 00:00:50,770
Government and Leader of the Mind and Behaviour Research Group at the University of Oxford.

10
00:00:51,490 --> 00:00:55,210
Kate, thank you for joining me. Thank you so much for having me.

11
00:00:55,840 --> 00:01:00,579
It's great to be able to talk about research that seems to have actually had the

12
00:01:00,580 --> 00:01:05,530
impact rather than aspirational research that hopes maybe to have some impact.

13
00:01:05,530 --> 00:01:16,629
So it's great also for us to do this because, you know, as the centre, it's something we try to to achieve, to actually just do research,

14
00:01:16,630 --> 00:01:22,420
not just as relevant for academic purposes, which is a virtuous purpose in itself,

15
00:01:22,420 --> 00:01:26,600
but also to try to get further and see whether we can actually make a difference.

16
00:01:26,610 --> 00:01:31,959
So I would be delighted to hear a little bit more from you about, you know,

17
00:01:31,960 --> 00:01:36,860
what was what was the problem that your contribution tried to resolve here.

18
00:01:36,880 --> 00:01:41,560
Tell us a little bit more about what was going on then what you are trying to solve.

19
00:01:42,280 --> 00:01:46,350
So I'll take you back to the very early days of the COVID pandemic.

20
00:01:46,360 --> 00:01:49,870
And so this this impacts of what happened in South Africa.

21
00:01:50,500 --> 00:01:53,680
But the situation was very similar all across the developing world.

22
00:01:53,830 --> 00:01:59,110
So in in April 2020, there were about one in five South Africans going to bed hungry.

23
00:01:59,470 --> 00:02:05,140
And in many countries, you know, that lockdown, very hard work is very casual.

24
00:02:05,680 --> 00:02:13,180
And so millions of workers were laid off. And if there was any government support, it has tended to be through food parcel distribution.

25
00:02:13,540 --> 00:02:19,480
And so in South Africa, like in many places, the existing system of getting aid out was just completely overwhelmed.

26
00:02:19,660 --> 00:02:22,560
There were issues with logistical management, there was theft.

27
00:02:22,900 --> 00:02:28,870
And so you just saw these nightly images on the news of these long queues for food and they were all,

28
00:02:29,050 --> 00:02:34,720
you know, running out and that that really dire situation was happening across the developing world.

29
00:02:35,560 --> 00:02:43,660
We know these problems are quite predictable. So there's actually been a long standing evidence based in development economics over the last

30
00:02:43,660 --> 00:02:50,470
20 years that tried to change what we what economists now think of as quite an ideological,

31
00:02:50,710 --> 00:02:56,830
non evidence based approach to aid, which is that it has to go out as food policy makers worry a lot,

32
00:02:56,830 --> 00:03:00,490
that if you just give recipients cash, they'll spend it frivolously.

33
00:03:01,270 --> 00:03:07,420
And so they either give us food or if welfare is given as cash like to jobseekers,

34
00:03:07,570 --> 00:03:13,300
it often has these quite strict conditions attached like you have to you have to be searching for a job.

35
00:03:13,720 --> 00:03:23,030
And so in this policy situation, the existing research base that we had, which are very strong, is quite a radical new form of welfare.

36
00:03:23,050 --> 00:03:30,430
It's the idea that you can just give cash directly to poor households, including during emergencies and pre-COVID,

37
00:03:30,430 --> 00:03:37,210
you and I, as well as big team of other researchers Rob GARLICK, Marine Mahmood Richard or Maya Johan House offer.

38
00:03:37,660 --> 00:03:41,559
We ran this big randomised trial in Kenya with GiveDirectly,

39
00:03:41,560 --> 00:03:49,540
this NGO that has really pioneered this approach to test what happens when you just give cash directly to poor households.

40
00:03:49,840 --> 00:03:55,720
And like a lot of other studies, we find that recipient households really use the money well quote unquote.

41
00:03:56,140 --> 00:04:00,550
You know, they spending it on food that improves children's nutrition and development.

42
00:04:00,790 --> 00:04:04,990
They also using cash to buy assets or businesses or to search for work.

43
00:04:05,950 --> 00:04:09,159
They often working, in fact, more rather than less.

44
00:04:09,160 --> 00:04:10,809
And they're certainly not wasting the money.

45
00:04:10,810 --> 00:04:16,720
They, you know, really investing it in things that can help them to improve their economic position in the future.

46
00:04:17,170 --> 00:04:22,720
And so what this research was trying to do was apply that evidence based, which we had pre-COVID,

47
00:04:22,990 --> 00:04:31,720
to the crisis of aid distribution that was happening during the COVID pandemic and say, look, actually, we know a shift that needs to happen.

48
00:04:32,260 --> 00:04:39,280
It's a really big shift. But we we need to stop trying to think we can get food parcels out and that that's going to solve this emergency.

49
00:04:39,430 --> 00:04:42,520
We need to do something that the policymakers may not approve of.

50
00:04:42,700 --> 00:04:48,370
But if we can give aid as cash, we'll be able to get it out much faster and to many more people.

51
00:04:49,030 --> 00:04:52,270
And that's that's really interesting because if we think about it, you know,

52
00:04:52,430 --> 00:04:57,970
the evidence base and as you also allude to around, you know, that cash can be really effective.

53
00:04:58,540 --> 00:05:02,110
It's been there now for a while. And it's it's an interesting.

54
00:05:02,440 --> 00:05:11,980
That every time, again, when you try to to do some of these things and you tell policymakers, well, we have the evidence, you can just do cash.

55
00:05:12,340 --> 00:05:16,760
I know from experience as well, a lot of people will have this prejudices.

56
00:05:17,020 --> 00:05:20,139
And and it's not just, you know, governments. It's middle classes.

57
00:05:20,140 --> 00:05:26,800
There's all kinds of people have these prejudices that actually giving cash will have these not beneficial impacts.

58
00:05:27,160 --> 00:05:33,430
So so it's one thing to supply the evidence, but clearly needs other work as well.

59
00:05:33,460 --> 00:05:40,930
So how did you go about convincing them? And I think that's the more almost the more important part, because we can keep on writing these papers,

60
00:05:41,080 --> 00:05:48,040
We can write a little block, we can say this, we have the evidence that we've seen so many countries that it's not been picked up.

61
00:05:48,040 --> 00:05:54,030
So tell us a little bit more how they came about and and how did you go about this?

62
00:05:54,870 --> 00:06:02,280
So I think the most important thing is that this could happen because there was this huge evidence.

63
00:06:02,910 --> 00:06:07,190
Researchers tend to get excited about the new study and their working paper.

64
00:06:07,200 --> 00:06:08,519
It's not even published yet.

65
00:06:08,520 --> 00:06:14,820
And then they want to go and say to the government, Oh, you should do this new thing at the frontier that I've come up with.

66
00:06:15,240 --> 00:06:20,910
And I think the key thing here was that we were able to say, you know, there's been systematic reviews.

67
00:06:20,910 --> 00:06:26,760
They know 160 studies all around the world in many different contexts that have similar findings.

68
00:06:26,920 --> 00:06:33,330
You know, we've done it recently in a context like Kenya, which is very similar to this country, but actually the evidence base was big.

69
00:06:33,610 --> 00:06:41,040
I think for me, that was that was important. In fact, possibly governments shouldn't be scaling one study.

70
00:06:41,040 --> 00:06:48,630
They should be scaling things where we have meta analysis, where we know that the evidence base is deep and it's uncontroversial what to do.

71
00:06:49,380 --> 00:06:56,160
So I think that was the first thing. It was really important when talking to government that we weren't just shopping for solution,

72
00:06:56,310 --> 00:07:05,430
we were doing evidence reviews across the broad range of the evidence across multiple contexts, including contexts that were similar to South Africa.

73
00:07:05,970 --> 00:07:12,720
To draw the conclusion. So I think that was the first thing was, you know, the evidence base is actually going to be quite deep.

74
00:07:13,380 --> 00:07:22,590
The second thing was, you know, changing format from the what a research paper was into what was actually going to be useful in the policy process.

75
00:07:23,010 --> 00:07:25,440
So the way that the impact happened,

76
00:07:25,620 --> 00:07:32,610
we work collaboratively with the presidency and the Social Security Agency and the Department for Social Development in South Africa.

77
00:07:33,060 --> 00:07:37,770
And there were a group of civil servants who knew of this evidence base,

78
00:07:38,130 --> 00:07:47,760
and they formed a working group on poverty that wanted to see whether the existing grant system in the country could be used to,

79
00:07:47,940 --> 00:07:56,040
you know, put in place these solutions. But the really important thing was workshopping the policy questions that they needed answers to,

80
00:07:56,050 --> 00:07:59,250
thinking about the formats that they needed the answers in.

81
00:07:59,490 --> 00:08:04,830
So it wasn't a research paper, it wasn't a policy brief. It was we're going to have to go through Cabinet resolutions.

82
00:08:04,830 --> 00:08:08,970
So we need a sort of short summary of of what's being proposed.

83
00:08:09,210 --> 00:08:16,170
And then we need the underpinning research work written in a really kind of clear, accessible format.

84
00:08:16,170 --> 00:08:21,840
And I think learning to do that writing was completely different from how you would write a research paper.

85
00:08:22,050 --> 00:08:28,860
You know, it's really focusing on what the evidence is across the range of papers rather than just one paper.

86
00:08:29,010 --> 00:08:36,600
So, you know, you're writing statements like the 20 studies in this field and six of them find this and, you know, 14 of them find that.

87
00:08:36,720 --> 00:08:42,810
So, you know, our best guess is that this is the this is what one should learn from that evidence.

88
00:08:43,050 --> 00:08:46,560
And I think that's really something economists don't do a lot in health.

89
00:08:46,560 --> 00:08:53,790
We do a lot of major studies, and we've actually worked a bit with the Mind and behaviour group on doing some some meta analysis studies.

90
00:08:54,030 --> 00:09:00,540
I think that was actually the most important skill that I was able to bring to this was sort of looking across the evidence.

91
00:09:00,840 --> 00:09:02,970
So I think that was the second thing.

92
00:09:03,210 --> 00:09:11,880
And then the third thing, which is something the CSA has been, you know, really worked at very hard, is having these deep,

93
00:09:11,880 --> 00:09:19,860
long collaborations with economists in developing countries so that you here we were working with the university in Cape Town.

94
00:09:19,860 --> 00:09:21,600
That was actually where I did my undergrad.

95
00:09:21,870 --> 00:09:31,740
So working with a team led by Ingrid Woolard, Mary Labour, and then we also had Maya Goldman to to Coachella a Jessica Nicola and Brandi Kraft.

96
00:09:31,920 --> 00:09:39,450
So it's a really deep team of both senior and more junior researchers who had huge government reputation.

97
00:09:39,660 --> 00:09:48,960
So they'd done consultation for the Treasury and the presidency before Mary was on the big panel that looked at social welfare even pre-pandemic.

98
00:09:49,290 --> 00:09:56,250
And so, you know, it was working with that team that we were able to achieve policy impact because we had local credibility.

99
00:09:56,490 --> 00:10:02,460
So I think those kind of three ingredients really were what brought together having the impact.

100
00:10:02,640 --> 00:10:05,610
But it's very different from what the current model is,

101
00:10:05,610 --> 00:10:11,610
particularly in the world of you do one child policy brief and then government is going to do the policy.

102
00:10:11,730 --> 00:10:15,600
This wasn't what it looked like at all, and I think that was a really important learning for us.

103
00:10:16,140 --> 00:10:19,890
So kind of push you a little bit on that terms of the way these debates would

104
00:10:19,890 --> 00:10:23,940
have gone and the way people were thinking about what is that evidence base,

105
00:10:23,940 --> 00:10:29,249
you know? You know, I don't know if any of the studies you could review referred specifically to South Africa,

106
00:10:29,250 --> 00:10:35,220
but often you get this experience that, yeah, maybe that may have worked in, I don't know, in in Ethiopia or in India.

107
00:10:35,230 --> 00:10:37,770
We want to work here. You know, we are different.

108
00:10:38,040 --> 00:10:43,919
You know, there is no external validity to this or surely the cost involved here would be so much higher.

109
00:10:43,920 --> 00:10:50,480
What are these numbers involved? So was there any attempts to do some modelling work this some some adjustment work?

110
00:10:50,490 --> 00:10:54,150
How how was that handled? Was that something the Treasury people were doing?

111
00:10:54,220 --> 00:10:58,440
Always other people doing. Tell me a bit more about that. And then how do you convince bureaucrats?

112
00:10:58,450 --> 00:11:03,219
How do you convince civil servants? I've been one, you know, it's not easy to convince them.

113
00:11:03,220 --> 00:11:06,550
So. So how do you think? Tell us a little bit more how that all went.

114
00:11:07,090 --> 00:11:13,550
I think first focusing on does the international evidence apply will not mean in this particular instance.

115
00:11:13,570 --> 00:11:13,959
Actually,

116
00:11:13,960 --> 00:11:23,170
the government was hugely eager to get that comparative evidence base because South Africa had this long history of doing cash grants for pensions.

117
00:11:23,380 --> 00:11:27,760
And there's a big child support grant that was put in place that's been enormously successful.

118
00:11:28,270 --> 00:11:34,239
So, you know, we didn't know about running cash grants and that that program is actually kind of world

119
00:11:34,240 --> 00:11:39,069
leading in terms of the number of beneficiaries who had managed manages to reach successfully.

120
00:11:39,070 --> 00:11:43,870
It's very well targeted, it's extremely pro-poor, and there's not a lot of leakage in the program.

121
00:11:43,870 --> 00:11:47,889
It's works on biometrics. So there was that success case.

122
00:11:47,890 --> 00:11:54,340
And I think, you know, had we been starting with the social welfare system from scratch, that probably would have would have been difficult.

123
00:11:54,850 --> 00:11:58,179
But they were incredibly eager to learn from other countries.

124
00:11:58,180 --> 00:12:04,900
So one of the most focussed on pieces in the policy brief we had set during the early days of the pandemic,

125
00:12:05,080 --> 00:12:10,090
we actually got a Venezuelan student who could read on what the Latin American countries were doing,

126
00:12:10,270 --> 00:12:14,740
you know, when it was in Spanish and Portuguese, you know, we couldn't even get it in English.

127
00:12:14,950 --> 00:12:20,079
But we had this comparative table of what are the different countries doing at the World Bank?

128
00:12:20,080 --> 00:12:26,920
So Hugo Giardini and that sort of social protection group, they were talking a lot about the evolving response to the crisis.

129
00:12:27,250 --> 00:12:34,360
And policymakers really wanted that because I think one of the things they were worried about was, you know,

130
00:12:35,170 --> 00:12:42,070
are investors going to think we're being completely profligate, just giving out money to our populations?

131
00:12:42,340 --> 00:12:48,129
And the fact that there was this broad movement towards increasing social protection during the pandemic in

132
00:12:48,130 --> 00:12:53,980
other countries was really important to reassure them that perhaps they wouldn't be that response in this case.

133
00:12:54,400 --> 00:12:58,420
So I think people were really eager to get the international experience.

134
00:12:58,660 --> 00:13:03,170
Actually, one of the things Marianne and Ingrid and team had worked on, you know,

135
00:13:03,220 --> 00:13:07,690
even in previous iterations, the work that they'd done was a deep mission,

136
00:13:07,690 --> 00:13:14,050
kind of talking to the people who did Bolsa Familia in Brazil and learning a bit more about how they did the targeting,

137
00:13:14,290 --> 00:13:19,240
because the approach that South Africa ended up using is the most similar to Brazil compared to other countries.

138
00:13:19,660 --> 00:13:22,809
So I think there was a real eagerness to learn from other countries.

139
00:13:22,810 --> 00:13:27,490
Yes. And not to sense that this wasn't relevant. And we have to have the evidence for our context.

140
00:13:27,850 --> 00:13:33,190
You know, policymakers are quite used to making the decisions based on not perfect information.

141
00:13:33,370 --> 00:13:40,570
I think sometimes we in especially in the world, we think, oh, we have to have done the audacity of this trial in this context at this time.

142
00:13:40,870 --> 00:13:46,750
And they were less worried about that. You sometimes having to make jumps, you know, and you don't have the perfect evidence.

143
00:13:47,020 --> 00:13:51,240
The other thing that you asked about cost that was super important and, you know,

144
00:13:51,260 --> 00:13:59,799
we did some cost calculations in our initial Kenya trial, but I didn't know anything about studying costs, which is a great shame.

145
00:13:59,800 --> 00:14:05,530
I think, you know, what the U.S. team brought. You know, they're not primarily ICT researchers.

146
00:14:05,530 --> 00:14:07,270
They weren't studying cash transfers.

147
00:14:07,390 --> 00:14:16,690
They had already built this model of the South African economy that one could use to study the poverty incidence of social transfers.

148
00:14:16,900 --> 00:14:23,200
And you know how if you also wanted to put in place tax increases, how that would affect levels of poverty.

149
00:14:23,530 --> 00:14:28,570
And so that wasn't a modelling technique I was familiar with as someone who primarily does field experiments,

150
00:14:28,810 --> 00:14:35,200
but that was the core of the work was being able to say this is how many poor people they are, you know, people below the poverty line.

151
00:14:35,320 --> 00:14:42,520
If you give this amount of money, this is how it's going to change what people's consumption levels are, whether they are living in poverty.

152
00:14:42,760 --> 00:14:49,390
And then if you do different versions of the grant, this is how many people are going to be eligible and what the poverty impact is going to be.

153
00:14:49,540 --> 00:14:56,170
And so that model was the core of the work that we did, and it was a real learning to me as an ICC researcher.

154
00:14:56,560 --> 00:14:59,140
You know, the benefits is only the first step.

155
00:14:59,350 --> 00:15:03,730
And then you have to be able to say to the government, who is this going to reach and how much is it going to cost?

156
00:15:03,970 --> 00:15:08,260
And those are not tools. We've even, you know, we just cost what's in the act.

157
00:15:08,440 --> 00:15:10,570
But that's it's not even the right question.

158
00:15:10,810 --> 00:15:17,650
So I think that really blew open for me what the discipline needs to be doing if you're really going to get national governments to scale.

159
00:15:17,860 --> 00:15:24,429
And I think it is collaborating much more with researchers in the sort of fiscal and public space to say,

160
00:15:24,430 --> 00:15:27,700
you know, how is this policy going to play out at an at a national scale?

161
00:15:28,210 --> 00:15:33,250
That's really interesting. Clearly, you know, it's one thing for a government to take up a policy and to do it.

162
00:15:33,250 --> 00:15:39,280
Let me gently suggest you wouldn't get an impact prize if there is no evidence that actually it

163
00:15:39,280 --> 00:15:44,560
had impact and that actually there is some evidence for South Africa of of the impact it have.

164
00:15:44,800 --> 00:15:49,450
You know, that there's there's one thing that I don't think I've ever told you really is that

165
00:15:49,450 --> 00:15:53,860
it was a peer review or actually over review on on the impact on poverty across the.

166
00:15:54,240 --> 00:15:57,290
From the COVID pandemic run by the World Bank.

167
00:15:57,710 --> 00:16:02,480
And it was really interesting because they just crunched the numbers from all of the world, whatever they can come in.

168
00:16:02,750 --> 00:16:08,690
And they kept on being worried because South Africa was an outlier. They said there must be something wrong with the numbers here and so on.

169
00:16:09,080 --> 00:16:15,709
And actually only in that process we could see and maybe there's one or two other countries that actually really

170
00:16:15,710 --> 00:16:22,940
have managed to to have a substantial poverty impact during the pandemic in ways that they were quite surprised by.

171
00:16:22,940 --> 00:16:30,379
And so so they ended up digging deeper. And and it is the case that seeing from the World Bank clearly from their comparative evidence that South

172
00:16:30,380 --> 00:16:36,980
Africa ended up being judged as one of the countries that better handled the poverty impacts than than many,

173
00:16:36,980 --> 00:16:42,040
if not most countries across the world. Now, I presume you will have other evidence.

174
00:16:42,050 --> 00:16:44,379
I'm sure the government has been trying to collect it.

175
00:16:44,380 --> 00:16:52,400
Tell us a little bit more what is the impact on beneficiaries and what kind of things have we learned since the basic policy shift?

176
00:16:52,670 --> 00:16:59,120
So we had the situation where they were giving I think they were giving out about 1.2 million food parcels a week,

177
00:16:59,480 --> 00:17:03,740
but they were about 10 million people who were below the breadline.

178
00:17:03,740 --> 00:17:08,780
So below the food poverty line, they didn't have enough income that they could actually get enough to eat.

179
00:17:09,170 --> 00:17:14,960
You know, the need was about a 10th of what was going to be necessary to avoid widespread hunger

180
00:17:15,470 --> 00:17:20,510
and what the government did in that situation instead of sticking with the food parcels.

181
00:17:20,510 --> 00:17:26,120
So the food parcels were actually constitutionally set up approach to social welfare.

182
00:17:26,120 --> 00:17:30,919
So, you know, like many developing countries, they didn't have a basic grant.

183
00:17:30,920 --> 00:17:36,230
If you were an adult, you were able bodied, so you didn't have a disability, but you were unemployed.

184
00:17:36,650 --> 00:17:39,650
And so the only thing that was there was these food parcels.

185
00:17:39,650 --> 00:17:44,570
If you were really in very dire need and you could go to your municipality and apply for them.

186
00:17:45,140 --> 00:17:50,000
So the shift that you know, with the city researchers and our team at Oxford,

187
00:17:50,120 --> 00:17:57,230
the shift that we achieved was instead of delivering aid through the food parcel system, the government did two things.

188
00:17:57,530 --> 00:18:02,870
They first increased the amount that was paid to people who are already getting a cash grant.

189
00:18:03,080 --> 00:18:11,270
So pensioners, women largely who were receiving child grants on behalf of their children and people who had a disability payment.

190
00:18:11,270 --> 00:18:14,750
So they temporarily increased the amount of the cash grants.

191
00:18:15,140 --> 00:18:19,700
We know already from the evidence that people share that money with their families.

192
00:18:19,700 --> 00:18:28,210
So the idea was, you know, already through the grant system, just by increasing the amount that flows out, you'll be able to reach more people.

193
00:18:28,640 --> 00:18:32,360
And then the second thing, which was the really, you know, remarkable policy shift,

194
00:18:32,360 --> 00:18:38,930
was to do this new monthly cash grant for 10 million able bodied unemployed people.

195
00:18:39,380 --> 00:18:47,570
They had previously not had any welfare payments, but the government would put in place an entire new grant in a six week period.

196
00:18:47,780 --> 00:18:49,460
And the technology was amazing.

197
00:18:49,580 --> 00:18:58,580
People mainly signed up for it for WhatsApp via WhatsApp, but they could also sign up online and so they signed up for the grant.

198
00:18:58,790 --> 00:19:01,610
Then you had checked, you know, this was the Brazilian approach.

199
00:19:01,700 --> 00:19:09,680
They checked against a set of data that, you know, a few things that they could check in Edmond data that you weren't very wealthy,

200
00:19:10,310 --> 00:19:15,980
you know, you didn't have a vehicle, you weren't getting unemployment insurance, you weren't on the government payroll, for example.

201
00:19:16,790 --> 00:19:22,160
And then later we actually built in that they looked at people's banking data to see their income,

202
00:19:22,730 --> 00:19:32,030
and then that that new technology meant that they were able to get eventually about 10 million people onto a new monthly cash grant.

203
00:19:32,300 --> 00:19:37,800
And so all in all, that meant that we went to reaching from, you know,

204
00:19:37,850 --> 00:19:45,230
1.2 million people who were getting the food parcels to 28 million people who were getting some sort of payment from the cash grant.

205
00:19:46,130 --> 00:19:50,090
So that was a huge shift in social welfare spending.

206
00:19:50,750 --> 00:19:57,170
In total, governments now spent about £7 billion, so that's about 3% of annual GDP.

207
00:19:57,170 --> 00:20:04,250
It's about the same scale as the as the UK furlough scheme through that kind of increase in in cash grants.

208
00:20:04,520 --> 00:20:10,880
So that's and that the research has shown that that's really been targeted very well at the country's poorest households.

209
00:20:11,720 --> 00:20:20,870
So the sort of set of testing means testing that they did has actually managed to have a really sort of poverty targeted grant.

210
00:20:21,650 --> 00:20:27,740
And so now they aren't they're no longer doing the increased child and child grants and pension payments,

211
00:20:27,980 --> 00:20:31,220
but that grant for the unemployed has stayed in place.

212
00:20:31,550 --> 00:20:40,010
And so we've we've modelled that with the team and we estimate that that's currently keeping about 2 million people out of severe poverty every month.

213
00:20:40,310 --> 00:20:43,550
So it's been a really big new kind of policy shift.

214
00:20:43,820 --> 00:20:47,300
It's the first grant for the unemployed in Africa.

215
00:20:47,330 --> 00:20:53,060
So it's it's been a really big extension of the social welfare system and it's really having quite.

216
00:20:53,140 --> 00:20:57,730
Remarkable effects on poverty. That's that's really fascinating.

217
00:20:57,820 --> 00:21:01,209
Maybe kids. As a final question. We're all researchers.

218
00:21:01,210 --> 00:21:04,440
We all try to have impact at times.

219
00:21:04,600 --> 00:21:09,790
What have you learned about trying to get impact from research from this experience?

220
00:21:10,580 --> 00:21:14,780
I think the the most important thing in this was was empathy.

221
00:21:15,080 --> 00:21:23,130
It was about stepping out of your academic world and trying to put yourself in the shoes of the civil servants who were trying to solve the problem.

222
00:21:23,180 --> 00:21:30,920
And so we worked really closely with this group. They were just incredibly committed to deliver that kind of impact.

223
00:21:31,100 --> 00:21:33,560
They were just working round the clock for years.

224
00:21:34,370 --> 00:21:40,970
And so I think so to really say, if I was sitting facing this problem, what are the technical inputs that I would need?

225
00:21:41,300 --> 00:21:45,350
And you have to take yourself as a researcher out of the process.

226
00:21:45,530 --> 00:21:50,000
It's not about whether it's your particular thing that's got your name on it.

227
00:21:50,630 --> 00:21:58,700
You know, it's not about whether it's your particular solution. It's about what, you know, what's actually going to be the best in the situation.

228
00:21:58,970 --> 00:22:05,990
And I think drawing on the really broad range of evidence, looking at what, you know, other people in other countries have done,

229
00:22:06,620 --> 00:22:12,050
and then really thinking about what others, the civil servants worried about, Why are they worried about that?

230
00:22:12,240 --> 00:22:17,180
You know, what are the barriers to them, you know, taking this this policy to scale?

231
00:22:17,510 --> 00:22:20,750
Because I think that's thinking partly about the technical challenges.

232
00:22:20,960 --> 00:22:24,860
And, you know, they were really deep in the weeds of the numbers and the model,

233
00:22:25,460 --> 00:22:29,930
but trying to make the model so that it answered questions that they had.

234
00:22:30,440 --> 00:22:34,310
But then also thinking about the politics, like how are you going to get this grant through?

235
00:22:34,520 --> 00:22:40,520
And so I think this for us, there's this huge worry that the cash grant builds dependency and that's a you know,

236
00:22:40,520 --> 00:22:43,850
that that has been there in wealthy states for hundreds of years.

237
00:22:44,180 --> 00:22:49,399
But actually to have this evidence base, we don't talk about it emotionally, but it's to say, you know,

238
00:22:49,400 --> 00:22:54,620
giving people cash builds resilience, it builds autonomy, it builds people's ability to make decisions.

239
00:22:54,620 --> 00:23:00,980
And, you know, they are the owners of the economic lives. And, you know, you're giving them the most flexible tool that you can.

240
00:23:01,400 --> 00:23:05,060
And that narrative of empowerment was hugely politically important.

241
00:23:05,300 --> 00:23:11,180
And so once we realised, you know, wasn't a lot of the evidence is saying just your food, calorie consumption improves.

242
00:23:11,180 --> 00:23:18,200
But the thing that actually gripped people was people can look for work, they can buy assets and that they weren't reviews on that so much.

243
00:23:18,200 --> 00:23:22,640
And we really focussed on that evidence base and I think that was actually what swung it.

244
00:23:22,880 --> 00:23:28,670
And that's also been what's kept the grant in place as part of the economic recovery is,

245
00:23:28,670 --> 00:23:36,200
is saying this is a tool to help, you know, where the state is struggling to get that sort of economic growth going.

246
00:23:36,350 --> 00:23:40,790
You actually putting this in the hands of the citizens and those are very effective economic agents.

247
00:23:41,180 --> 00:23:46,520
So I think that change in narrative was actually the most the most important thing, and that came from China.

248
00:23:46,520 --> 00:23:50,220
Understand the political reality that people were facing. Yeah.

249
00:23:50,240 --> 00:23:55,100
So I think empathy, empathy with your policymakers is my my catch phrase.

250
00:23:55,910 --> 00:24:00,350
Very good. Well, it will come. Thank you very much for talking to us.

251
00:24:00,350 --> 00:24:08,780
And also, again, congratulations for winning the award of outstanding public policy Impact of the Economic and Social Research Council of the UK.

252
00:24:08,930 --> 00:24:10,190
Thank you very much for talking us.