 So thank you so much for having me and thank you all for coming to the session. So I'll present a review paper, which is a bit unusual, on the impact of inequality on growth, human development and governance. It's a joint paper with Rachel and Finn, who are here. And it's part of the project financed by Novornoordisk. And as Rachel explained, this is an exercise that we did right in the beginning of the project to try to take stock and understand what do we know and where are the gaps. So our motivation is exactly the motivation of this conference. Inequality is a great challenge, but we also need clear evidence for an informed policy debate. So here we try to take stock of the current knowledge on the effects, on three outcomes that are important for development. So we focus on economic growth, on human development, and in particularly on health and education, and finally, governance focusing on democracy. We try to understand what are the core arguments and also the underlying mechanisms. And after focusing on the theory, we go to the empirical evidence. In this paper, we focus mainly on cross-country insights, and you'll see that reflected in the presentation as well. And we are mainly looking at what has been done in economics, but we draw some insights from other disciplines such as political science or some health studies. So I'll start by giving you our main insights. And we try to first bring some clarity through a general framework. So a simplified framework of the theory, what are the main arguments and what have been the transmission channels identified. We focus mainly on these channels separately, but we also draw attention on the linkages between them. We then move to the evidence. And our main conclusion is that there's no clear consensus emerging from the evidence, and there's a need for further in-depth studies on specific transmission channels in particular. And then we kind of summarize some of the main empirical challenges that have been identified. And this is the structure that I'll follow for the presentation today. So I'll give you a bird's eye view on the theory. Then I'll move on to the main insights from the empirical evidence, and I'll summarize some empirical challenges before I conclude. So starting with the theory, and apologies if you can't see very well at the back, we kind of tried to summarize everything into a simplified framework. So it is simplified and we're losing the details here, but we try to put together something that highlights both the differences between the different channels, but also how they're interlinked. So if we start from the left to the right of the framework, we have some transmission channels, the intermediate effects, and on the right you'll see the general positive or negative effect on the three main outcomes of interest. Then from the top to the bottom, we try to think about the underlying drivers. So broadly speaking, we divide them into the effects through the poor, through the population at large, or the average effects, and the wealthy. So I will not give you all the detail, but bear with me while I highlight some of the main transmission channels. So we see first that in general there's more indication of negative effects on these outcomes, which is unlike what initially had been proposed by classical theorists. So they had highlighted that inequality through a higher level of income would have a positive effect on growth, through the incentives on the one hand, but also a higher marginal propensity to save. So richer people would, they said, have a higher marginal propensity to save, and that would have a positive impact on growth. Then we see that the theoretical insights after that kind of draw attention to negative effects on different outcomes. So for example, it has been said that inequality discourages growth through the discouragement of investment via weak property rights and also the regulatory framework. So here we're talking about the effect of inequality on policymaking and on political instability. This has a negative effect on growth, but also on democracy and governance through the negative effect on lower institutional quality. We can also think about spending in public goods. So if we focus on the reach and if inequality leads to favoring private over public investment, this leads to lower spending on public health and education, with a direct effect on education and on growth. There are some political economy theorists that focused on the median voter theorem. This predicts that taxation is a percentage of income, and a lower well endowed median voter will tend to have a preference for a higher equilibrium level of taxation with a negative effect on growth. If we focus also on the small middle class, there were some studies suggesting that the effect through demand on manufacturers will have a negative effect on growth, but there's also some political science literature on the effects on the demand for democracy. I'll jump back down and focus on the poor. So again, political science work focusing on the effect of polarization and power concentration and how that affects demand for democracy. So focusing on the disconnect between the demands of the rich and the poor, but also in terms of political violence and trust, linking to the effect also on health at the individual level. And finally, there were some studies focusing on what happens when there's constraints in terms of access to credit and how that can lead to underinvestment in human capital. So here the demand for education, poor families may tend to prefer larger families as well, so they focus on high fertility. And both these mechanisms highlight how the situation of the poor in a context of high inequality can lead to underinvestment in human capital. Everyone has an effect on human development, but also on growth. So hopefully you will have seen that some of these channels highlight negative effects on the three outcomes, as well as on specific ones. So moving on to the evidence, I don't want you to see all of the references, and I'll try not to bring a lot of specific work, but give you more of a general view. So focus first on the evidence on growth, and this is mainly work focusing on reduced-bore equations. So the general viral style growth equations, you have economic growth as the dependent variable, and then inequality is added in as one independent variable, so focusing on the coefficient. So early work you see at the top, mainly used cross-section data. So data availability has been a challenge from the beginning, and mostly they found a negative effect of inequality on growth. When the Dininger and Squire data set appeared, there was the possibility to do panel analysis. This opened up in terms of the methods that could be used, and there were some studies suggesting that actually growth had a positive effect, sorry, inequality had a positive effect on growth. At the same time, some started questioning and saying, well, actually the effect depends. So Barrow has a paper in 2000 saying that the effect depends on the level of development in the country, and then Banerjee and Duflo also started questioning the functional form of the equation and saying, actually there's a non-linear effect. More recent work, and also with the emergence of the new data from the wheat and other smaller attempts at providing more data appearing, the kind of results in the more recent literature mirror the disagreement in the early work. So some studies find a negative effect, some suggest that it depends, and a smaller group of studies find a positive effect. So if I move on and try to tell you a little bit more about the channels of transmission and focusing on our three outcomes, so education and health under human development, the first line refers to the empirical literature on the effect of inequality on growth. I told you already that the reduced form studies give us some mixed evidence. Then there are some studies that suggest that indeed there's a positive effect via savings if we looked at micro data at the household level, but the evidence from cross-country is that we don't know. So mixed evidence on that. Studies suggest that indeed there's a negative effect through the credit market imperfection, so the lack of access to credit markets and the high fertility channels, but not so much that there's evidence on the political economy kind of arguments through government expenditures and taxation. Again, we don't know in terms of the structure of demand, the evidence is not conclusive, and studies suggest that indeed there's a negative effect through political instability and corruption. In particular, the more recent work has confirmed the evidence of some effects through credit market imperfections and high fertility. So moving on to education. The empirical literature is more on the effect of education on inequality rather than on inequality, so the opposite direction of causality. There's one study that actually finds that more expenditure on education is correlated with higher inequality, sorry, higher inequality is correlated with more expenditure on education, and some studies suggest that there's a negative effect of inequality on enrollment and attainment. In terms of health, when we look at the effects of inequality on the health of all the individuals, there seems to be a disagreement of the economics literature compared to the public health studies, whereas in the public health they say there is a negative effect of inequality on health. We can see that the economic studies caution and say actually the evidence is not there. I won't tell you much about this group of work. They have similar hypotheses that vary and focus on the relative position of individuals, so they're more at the individual level, and the evidence is mixed there. I should also say that we only found one paper discussing the impact of inequality on human development by David and co-authors, and he will tell us a little bit more in a bit. In terms of governance, again, mixed evidence on the effect of inequality on the stability and transition to democracy, some confirmation of a negative effect on institutional inequality and on political participation. So I promise to summarize some of the main problems that have been highlighted in terms of the challenges of establishing this effect. I'll start with data quality and availability. There has been, of course, great progress achieved, but there is still some discussion of the problems when there's lack of consistency and the definitions used on the sources of data and how data is processed to the end result when researchers can access the latest forms. So how data is treated, for example, pretext, text, posttext, and all of that. So, of course, this creates empirical challenges. There's specialty scarcity when we need to test these mechanisms at the individual level. Concept and measurement of inequality is, of course, also consequential. Different specific concepts will be used to test different and specific mechanisms and the measures will also have an effect on the weight that is given to different parts of the income distribution. The choice of indicators can have implications for the results and there have been some specific criticism towards relative measures. So part of the idea of this project is to explore exactly what happens when we move away from relative measures of inequality towards thinking in absolute measures. I'll just say briefly that, of course, estimation methods can lead to different results. This happens when we compare the results from cross-sections to panel data estimations. Both of them can suffer from measurement error. I mentioned the study questioning the linear form of the specifications and despite many efforts to tackle the reverse causality, there's always some sort of questioning that we cannot really untangle the effects. So to summarize our main conclusions that have been helpful also to set up the work for the rest of the project are that evidence from reduced form equations is not consensual and that has been less attention in the literature on the transmission channels. There's a need for further research on education outcomes and there's no clear support for the negative effect.