Democracy Matters in Global Health

Panelists discuss the links between democratic governance and global health.

BOLLYKY: Great. We’re going to get started. Good afternoon. Thank you for coming on a beautiful spring day. My name is Tom Bollyky and I’m the director of the Global Health Program here at the Council on Foreign Relations. You’re here for a session on democracy and health. Thank you to those of you that have joined us online. If you haven’t already silenced your cellphones, I’d be grateful if you did so. Today’s event is on the record and it is being livestreamed. This will be an unusual format for a CFR roundtable. And I’m going to beg your indulgence with that.

I’m going to start by giving an overview of the paper that my colleagues and I published today in The Lancet. It’s the impetus for this meeting. It assesses the links between democracy, adult health, and disease-specific morality in 1970 countries over a forty-six-year period, between 1970 and 2016. We’re also releasing today a data interactive on CFR.org that’s based on those results. At the end of this overview, I’ve been asked to just turn it over to the first panel, rather than taking questions. So in the meantime, my colleagues will sit here awkwardly on stage while I do these initial remarks. They will then get the chance to do a panel for forty-five minutes on what the global trends are on democracy, on health, and on health spending, and what we know about them and how we know about them.

The speakers on this panel, Joe Dieleman and Simon Wigley, are coauthors on this paper. They’re also more than capable of answering your questions about that paper. My terrific colleague Alyssa Ayres is moderating the panel. You will see me again, however, so if you do have questions you can try to work them into the second panel, which I will be moderating. The second panel has two speakers who are not associated with this paper but are most definitely expert in the topics of democracy internationally and its promotion. Those speakers are Tom Carothers from the Carnegie Endowment and Kate Somvongsiri from USAID. They will speak, again, about democracy promotion internationally and ideas for extending its reach and its effectiveness.

I want to just make a couple of acknowledgements before we move on. The first is to the CFR meetings team to thank them for organizing this great event. And the second is to Bloomberg Philanthropies for their generous support for this research and this roundtable series. This research is very much a product of collaboration of several coauthors and institutions. And in particular, I want to acknowledge Tara Templin who is at Stanford and who led a lot of the analytical work underlying this paper, and Diana Schoder who’s here at CFR. And they’re both in the audience as well. I’m also contractually obliged to acknowledge the fact that the work you’re about to hear is based in part on my book that was released last fall. So it’s Plagues and the Paradox of Progress: Why the World is Getting Healthier in Worrisome Ways. And with that, let’s get to what we’re here to talk about today.

Democracy in recent decades has played what has seemed like a modest role in global health. For decades, Cuba and China were held out as countries that had provided good health to their populations at low cost. Many of the nations that have had the greatest gains in life expectancy—Ethiopia, Uganda, Myanmar, Rwanda—are countries that are non-democratic, or autocratic. They have achieved their successes often with the heavy support of international aid. But health needs are starting to change in low- and middle-income countries. So the child deaths worldwide over the last fifteen years have almost halved. Infectious diseases no longer cause the majority of death or disability in any region of the world, including sub-Saharan Africa.

And that leads to the fact that when you have these reductions in infectious diseases and, of course, you have these reductions in child deaths, and you save the lives of children, you’re going to get more adults. And that’s what we’re seeing. And adults have adult health needs. So you are getting more adults and you’re having populations grow. And that is happening at a rate that’s much—or, that effect is overwhelming the fact that, you know, rates of noncommunicable diseases, diseases that disproportionately affect adults—heart disease, diabetes, cancers—they’re not necessarily increasing the rate. But the population that is likely or susceptible to get those diseases, you’re seeing more of them because people are aging and you’re seeing population growth. And that’s really driving the fast rate of increase in these diseases.

You can see that from the share of health burdens in low and middle-income countries that are now represented by these noncommunicable diseases. Take an example: Bangladesh from 1990 will go from a country that had roughly 28 percent of its death and disability associated with these noncommunicable diseases. By 2040, that’ll be 82 percent. That’s roughly the same as it is in the United States, but they’ll do that transition in fifty years instead of two hundred. You see similar results with countries like Ethiopia and Myanmar, which are also having very dramatic shifts in health needs.

What that led us to question is whether or not the results, or the modest results we’ve seen from democracy in the past on life expectancy and on child deaths, whether that would also be true for these noncommunicable diseases or health concerns of adults. And what we find is, not so much. So the first analysis that we did was looking at countries that transitioned to democracy between 1970 and 2015, and compared their adult life expectancy, life expectancy at fifteen, with countries that had not transitioned during that time period. So it’s about fifteen countries that had transitioned, fifty-five countries that had not. And we controlled for HIV, because HIV is a disease that predominantly affects adults, but is heavily targeted by development assistance for health, international aid, and that’s true both in democracies and in autocracies. And so we didn’t want to get that confounding effect.

But what you see is for the countries that transitioned, their adult life expectancy increased by 3 percent after the first ten years of that transition. And that’s a pretty dramatic difference from other nations. You can see that here on your left-hand side. And the right-hand figure shows the probability that those results are the byproduct of chance. And that probability’s very low. All right. So the next analysis we did was to look at what drivers, what explains the mortality reductions you see from various diseases within a country. And we compared democracy, or democratic experience—which is the quality of the democracy and how long the country has been a democracy—to other factors like GDP per capita or urbanicity or aid.

And what you find is that democratic experience explains more of the reductions and is associated more strongly with more of the reductions in mortality from diseases like cardiovascular disease, transportation injuries—which are like road traffic injuries—cancers, and other noncommunicable diseases. They’re more associative than GDP. So for cardiovascular disease, it represents almost—close to a quarter, or explains as close to a quarter of those differences, and for the other causes a little bit less. You do not, however, see that many infectious diseases are associated with improvements with increases in democratic experience. You also see that the role of democracy is increasing over time. This shows you just for cardiovascular diseases, and increased from roughly 13 percent to, again, around 23 percent of explaining, again, the differences within countries from—in morality.

So what explains this outsized effect—health effect of democracy on noncommunicable diseases, on adult health versus infectious diseases that disproportionately affect children? Well, one thing we wanted to make sure is that we’re not just seeing a byproduct of democracies being wealthier. And what you see here is that increases from democracy are, in fact, not associated with increases in GDP. They are, however, associated—those increases—with increases in government health spending and declines in mortality from cardiovascular disease. How big of an effect is this having? Well, for each point increase in democratic stock, again, democratic experience, you see a 2 percent reduction in mortality from cardiovascular disease over twenty years. For cardiovascular diseases, transport injuries, tuberculosis and other noncommunicable diseases, it’s about a 9 percent total decrease in morality over twenty years.

That may not sound a lot, but the median country increased its democratic experience or its democratic stock by 4.88 points over those twenty years. So what that means, is if you just look at cardiovascular diseases alone, we estimate improvements in democracy, averting 16.2 million deaths globally. Because you see this dramatic shift of noncommunicable diseases to more populous countries, if you just extent that rate of improvement out to 2017 it goes to twenty-one and a half million. Again, because of this dramatic shift that’s happening.

So to try to make these results a little more policy actionable, we tried to identify which component of democracy is really driving these health benefits. And what you see is if you remove them one by one, it’s free and fair elections. So if you remove free and fair elections from the drivers or the overall components of democracy, the health effects cease to be statistically significant. They effectively disappear.

So what conclusions do we draw from these findings? We draw the fact that increasingly democracy or elections and the health of the people are inseparable. That this suggests that there needs to be more support for development agency programs that promote democracy in governance. And this will be particularly true as it seems unlikely we’re going to see dramatic increases in global health aid in the future, particularly for NCDs and injuries. That looking at democracy promotion and governance may be a way to see big health benefits without necessarily seeing large increases in development assistance.

With that, I will turn it over to the panel. (Applause.)

AYRES: Thank you. Thank you for that fascinating introduction to the topic here. Welcome to our first panel in today’s event on Democracy Matters in Global Health. We’ve got with us our two speakers, Simon Wigley and Joe Dieleman. I am—you have their bios with you in your reading packet. My name is Alyssa Ayres and I’m senior fellow here at the Council on Foreign Relations. I’m regionally focused. I focus on India, Pakistan, and South Asia, where you’ve got some big democracies and some health challenges and some health improvements. But I’m looking forward to hearing more from both Simon and Joe. I think we’re going to lead off with Simon.

WIGLEY: Sure, sure. So I’m going to focus more on the democracy side of the study but try and connect it to some of the results that Tom mentioned. And perhaps I could start with a general sort of statement about the argument that’s sitting in the background here. And it’s familiar to those who’ve read Amartya Sen’s Development as Freedom, where he argues that democracies have a wider—that government has a wider support base than an autocracy. So it has more—the political survival of democratic leaders depends on obtaining more votes, which means the need for spreading health care to a wider set of the population is higher. But also there’s an important feedback loop that democratic leaders can receive feedback about the need for intervention, and success of that intervention.

So this is kind of sitting in the background as the main argument. The strength of the study, a lot of it comes down to the strength of the data. We rely on two datasets, one of which Joe will explain a lot more detail, and better than I can, from the IHME and Global Burden of Disease database. The other one relies on a relatively new study on indicators of democracy, called the Varieties of Democracy dataset, which is based in the University of Gothenburg. And this is a step up from the usual indicators that we’ve been using in political science and international relations, which is Freedom House and, for example, Polity IV. It’s a step up because of the level of sophistication in dealing with potential measurement error is significant.

The other key part of the V-dem dataset is that it has incredible detail. So rather than just looking at general components of democracy—suffrage, free and fair elections, free speech, et cetera—you can dive a lot deeper and look at specific signals of the health—the democratic health of a state. For example, whether the electoral commission is autonomous from the government or not, or whether the degree of vote-buying that exists in each state. So this is a really impressive dataset that has been put together. As Tom mentioned, we adjust this, we rely on core components of democracy—free and fair, suffrage, freedom of association, et cetera. But we also add in a historical component to capture not just the ebb and flow of democracy across time, but also how long a country has this level of democracy, let’s say. So let’s more capturing the kind of long-lasting quality of democracy in each state, rather than day-to-day fluctuations.

So perhaps I could briefly mention some of the trends in democracy that have emerged that we can observe from this new dataset. So the good news is that fifteen countries transitioned to democracy between 1980 and 2000, as Tom mentioned. And we use this as part of—one of our pieces of analysis. And most countries have improved in terms of their democratic experience, or democratic quality over time, since 1990. So there’s the positive. And that is quite a big positive. Most countries in the world are democratic right now. The worrying news is familiar to us already perhaps. Since 1970, fifty-five countries have been entirely autocratic without interruption.

And there’s growing concern about the democratic erosion in some states in the last two to three years. There’s the familiar cases, such as Russia or Venezuela, but also recently in the last two years, perhaps even some more established democracies like Brazil and India seem to be at risk of some democratic erosion. One of the telltale signals for this might be from the V-dem dataset is the autonomy of the electoral board, how much interference the government has in terms of the electoral commission. And you see in those states that it’s actually degraded the last two to three years, which is kind of telltale signal of a risk of further erosion, perhaps.

As Tom mentioned, free and fair elections is what stands out for our study. And this connects directly with this democratic erosion I just referred to. So those countries that are eroding, they tend to fit the category of what is now called electoral authoritarianism, or competitive autocracy, where you have the formal conditions for democracy—regular elections, opposition parties, perhaps even multiple media outlets—and yet, the incumbent can stack the playing field to such a degree that the chances that they will be elected out of office are very, very low.

So they can create the façade of democracy, but in fact the idea of rotating leaders is—empirically usually doesn’t happen. And the free and fair elections is the—the façade is—disappears at this point, because if you control the media, even if there’s diverse media outlets, if you control the electoral board, if you can use vote buying in poorer neighborhoods, effectively you can really determine the outcome of the elections to a large degree. And this one element, as I said, is the key part of our study, in a sense, the part which directly connects with adult health and noncommunicable diseases.

AYRES: Joe.

DIELEMAN: Thanks, Simon.

So if you glanced at the bio you would have seen that I work at the Institute for Health Metrics and Evaluation, which is in Seattle, Washington—part of the University of Washington. And I highlight that only to say that that’s kind of my expertise. That’s what I think, in many ways, I can bring to the studies, is thinking how we measure health and health outcomes. And so in the few minutes of remarks I have here, I’m going to really focus on a couple characteristics of how we measure health. I’ll then kind of dovetail and talk very briefly about some of my own research as a health economist. And then I’ll kind of tie it all together, hopefully.

So just starting with health data, I tend to think that if you’re thinking about mortality statistics or just health data in general, there’s kind of two camps. The more traditional set of statistics are what we who do the statistics and the data collection call all-cause statistics. And that simply a metric like life expectancy or child mortality. And those statistics are very valuable. They can be compared across time. They tell you a lot about an individual country, or a state, or a city, wherever it’s being measured.

But what they don’t tell you about is the actual cause of death. I think that’s where a lot of the advances have taken place over the last two or three decades is on a global level, but also a much more microlevel for states, and provinces, and even cities, measuring the causes of death. Not just the death rate, how many people die, but more specifically how many people died from ischemic heart disease or how many people died from diabetes. And I think that’s an important distinction to make. First of all, the methods that you measure those things are quite distinct. But from a more pragmatic policy perspective, I think cause-specific morality and cause-specific health outcome data is quite a bit more actionable.

It’s actionable, first and foremost, because you can think about the interventions needed to care for and treat the patients before mortality. It’s also actionable because you can start to think about prevention. Again, when it’s just all-cause mortality, it’s very hard to know what people are dying from, and hard to understand what maybe risk factors are leading to those deaths. On the contrary, when you’re looking at cause-specific mortality, you can start to back out and understand what are the risk factors? What are the conditions that are leading to mortality? So, for example, we look at a very large set of modifiable risk factors, those are things that we think can be changed, and measure essentially how much mortality globally is attributable to those risk factors.

When I say “risk factors” I mean things like you’re all familiar with and certainly see in the media and see in your own life, use of tobacco, diet, obesity, physical activity, air pollution. Those sorts of things—again, from a pragmatic policy perspective, you can start to think about interventions that prevent those risks and ultimately have an impact on changing mortality. So ultimately my thesis is that really cause-specific health estimates are of far more utility.

And that’s really what the institute that I work for has essentially gained a fair amount of notoriety for, is the measurement of the global burden of disease, which is the world’s—currently the world’s largest publishing scientific endeavor. It includes over eight thousand researchers that are scattered throughout the world that essentially measure everything I just said. Measure mortality rates, but not mortality that is all cause, but instead look at mortality for, like I said, specific diseases, diabetes, HIV, cardiovascular diseases, breast cancer, so on and so forth.

And really the last reason I want to point out—the last kind of reason I think cause-specific mortality and cause-specific health outcomes are of great utility, is because they can start to be connected to broader research endeavors or, in lines with my own work, it can be connected with health financing data. So I mentioned before that I’m a health economist by training. My research is really—revolved around research tracking, just understanding how much money is being spent on health. And as we evolve and kind of push forward this kind of discipline, that the steps that we’re taking is to measure health spending at a more granular level. And part of that means just identifying what health conditions resources are being spent on.

So for example, we have a research study in the United States that takes the $3 trillion spent on health care and just breaks that $3 trillion up into the diseases and the populations that that spending is coming from. And it, you know, comes to a wide variety of conclusions about where the money’s going. And it’s not always necessarily to the places that people think we’re spending the most on.

One of the long-standing research projects that I’m involved in is tracking development assistance for health. So development assistance for health is essentially international aid. It’s coming from, in most cases, high-income countries, some upper-middle-income countries as well. Oftentimes it goes through major development agencies, such as bilateral aid agencies, but also multilateral aid agencies, such as the World Bank, or the Global Fund, the U.N. system, and whatnot. And those resources are spent in low and middle-income countries with the intent of improving, or at least maintaining, health care.

And that project has tracked development assistance for health from 1990 through the present. And what we see over the course of that, you know, nearly thirty years is that there’s been a five-times increase in the amount of development assistance that’s provided. And in 2018, our most recent numbers—actually, not quite published yet, but soon to be published—we see that development assistance is just under $40 billion. And if you look at the growth trends over time, most of that increase in spending really was the first decade of the new millennium. From 2000 to 2010, the growth rates in the amount of development assistance for health that was provided really skyrocketed. And it was essentially in some ways a global commitment to global health, and in particular towards infectious disease, that led to a major increase in spending.

So I said our study starts in 1990. In 1990, over 50 percent of development assistance for health was for child health and maternal health. A lot of money on vaccines and a number of other interventions specifically for safe pregnancy, labor and delivery, and for those first few months, and ultimately years, of a child’s life. If you fast-forward to 2010, I told you that there’s been a major surge in the amount of development assistance for health that’s provided. And most of that surge focused on HIV, TB and malaria. It was the—essentially the genesis of a number of major organizations—such as the Global Fund and others—that provided resources focusing on those diseases. And in the U.S. government here was the PEPFAR program that really took off and provided literally billions of dollars of development assistance.

So where in 1990 over half of the development assistance was for maternal and child health, if you fast-forward to 2010 roughly 50 percent of the development assistance was for HIV and malaria alone. So you see a really major shift in the composition of the development assistance for health that was being provided by high-income countries. Since it relates to this topic I’ll highlight too that noncommunicable diseases, which we also track and also development assistance for health targets, is really relatively consistent, about 2 percent. So a very small fraction of development assistance for health targeting noncommunicable diseases.

So I said I’d talk briefly about health data, a little bit my own research, and then now to try to pull it all together. So the question is, why do these things matter and how, in my mind, are they connected? So two things come to mind right away. The first Tom already alluded to. He, without saying it, was talking about what academics will call the epidemiological transition. And so the epidemiological transition may sound a little fancy, but the idea is that the epidemiology is essentially: What are we dying from? And the transition is the fact that what we are dying from is changing over time.

And the hypothesis of that epidemiological transition was highlighted in some of the slides that Tom already showed. But it is that over time there’s movement away from infectious and communicable diseases towards noncommunicable diseases. And I think my point, as a a health scientist studying specifically metrics, is that you really need that cause-specific information to be able to test the hypothesis of the epidemiological transition. And it’s really only with cause-specific information that we understand what people are dying from.

The other thing that I think is really interesting about the epidemiological transition is that the rate of this transition has been changing. And I think that doubles back to some of the things that we talked about, or I mentioned earlier, related to development assistance for health and technology changes. And so in the past, where transitioning from high-mortality related to infectious diseases and moving to noncommunicable diseases may have taken a set amount of time, we see that more recently, especially in low and lower-middle income countries, that transition is happening much faster.

And I think the consequence, and the really important component of that, is that change, which moves faster, isn’t quite as dependent on economic development as it used to be. And it’s also not as dependent on the development of a domestic health system that is able to necessarily treat noncommunicable diseases. I think that lends itself, or at least leads to, a very concerning problem where there’s movement towards the primary causes of the noncommunicable diseases, but there’s not a health system necessarily that’s been developed well to take care of noncommunicable diseases.

The second point of how this all comes together is simply it allows us to study things, like we’ve done in this paper that was published today. Again, there’s a history of looking at the relationship of democracy and health outcomes. But really, I think that story becomes much more nuanced and much more policy-relevant if we can understand specifically which health outcomes are relevant to democracy. And in particular, some of the findings that—and conclusions that we came to—is that it’s these noncommunicable diseases, which, again, according to this transition are more and more prevalent, are increasingly a function of democracy and relevant to democracy.

So to close I’ll end with just three very salient points, three things to take away. Number one, I really believe infectious and communicable diseases are still a very big deal. Over ten million people die each year from communicable diseases. That said, noncommunicable diseases are an increasingly large component of the causes that lead to mortality and morbidity. Second point is that the health systems needed to treat those noncommunicable diseases are not self-evidence. And in many low and lower-middle income countries, those health systems have yet to be developed.

And then the third conclusion is that you really need health metrics that are at the cause level that are breaking out mortality, morbidity, and health financing for specific diseases in order to be able to track and understand these trends across time. With that, I’ll conclude.

AYRES: Thank you. So I want to ask just a couple questions to get us going before we open it up to members here. And I’m hoping to actually pull more out on the issue of the linkage between the findings of democracy and its effect on developing health systems that can adequately treat noncommunicable diseases. One of the things actually that struck me about the Lancet paper was that there is an important component of transportation injuries. This is not necessarily a—this struck me as more of a governance issue, not so much a health systems issue. But it seemed that the most important causes of death that were in this study were cardiovascular disease and transport injuries. And it goes on down, but can you speak about this issue of democracy and its linkages to developing health systems? Either Simon or Joe. Who wants to take that?

WIGLEY: You’re better on transport injuries, but I think it was a direct. So we—you have to correct me if I’m wrong. So we had—one of the pieces of analysis, whether there was a direct impact of democratic government or an indirect impact via spending or economic growth. And if I remember correctly, the transport injuries was a direct result of policy—speed limits, seat belts, drunk driving, et cetera. So if that’s correct, then, it’s a government—democratic governments being more interventionist on this level directly, rather than through spending, perhaps. But perhaps Joe has a different take on this.

DIELEMAN: Yeah, so, I mean, I think you can think of curbing mortality and health burden, morbidity, associated with traffic injuries in two ways. One, you can prevent those injuries from happening in the first place, and you can care more efficiently for people after they’ve been hurt. And I think the hypothesis here is that democracy certainly can impact both of those things. Democracy can impact prevention through many of the mechanisms that Simon just mentioned, and again through treatment—essentially through health-care system development and through health-care spending, which comes also with democracy—or, at least is encouraged by democracy—we see that a health system would be more efficient and effective at dealing with patients after they’ve been injured.

AYRES: Can you speak a little about the feedback loop, which is an important part of the argument that this article makes, that democracy provides a feedback loop for voters to either reelect or kick out governments in power. Do you see that as necessarily requiring voters to make specific demands about health delivery, service delivery, health-care systems? Or is it more a question of a kind of general sense of whether a government is being responsive to helping improve lives.

WIGLEY: So the feedback loop is just the signal through the voters or through civil society mobilization that’s a—and through the media. So if there’s a disease problem or a disease burden in a certain part of the population, democracies have the greater means to receive that signal, and to realize that policy intervention is required. But also probably equally importantly, to check whether the intervention is working. So especially in a closed autocracy, there’s no kind of—there’s no such feedback mechanism. And where free speech is limited, it becomes equally even more problematic.

So some have argued that in the case of China, they recognize this as a distinct problem. So the so-called firewall on social media is actually designed—and this is based on research in Harvard—that it’s designed not to stop complaints, so to speak, about bureaucrats, but to stop mobilization. So the filtering that’s going on in terms of social media is about not preventing complaints or the signals that’s something wrong, but to prevent collective action. So the Chinese government, if that’s correct, is recognizing the shortfall of autocratic or less than democratic states.

AYRES: Joe?

DIELEMAN: No, I don’t have much to add. I think Simon and maybe the second panel are probably better at thinking about that feedback.

AYRES: When you worked to develop the database and to match the database with health systems and health outcomes findings, can you speak a little bit about some observations from some of the countries that have gone through a democratic transition and experienced these improvements in health outcomes? Do you have some cases that—the Lancet article is very theoretical. It doesn’t give you a few of the cases that would be helpful to understanding how exactly this works.

WIGLEY: So the kind of stand-out examples of significant improvement after transition—so, you do have two kinds of analysis in the background, one in which we measure democracy in terms of the transition and sustained transition to democracy. The other one is this democratic experience indicator. The first one is a good way to pick out some cases, because the model compares a country with itself, so to speak. So for example, Poland or Brazil, when compared with themselves, have done remarkably well since the transition. And Bolivia as well. So there’s a few standout cases when you look at the curve or the gap between what would have been the case in Poland versus the doppelganger Poland.

You know, this is roughly—you know, there’s more than the 3 percent improvement in HIV-free adult life-expectancy. Because the 3 percent we were mentioning is the average across all these states. So there’s obviously variation. So this—you know, the average equates to roughly two years of life expectancy after ten years. So I think in the Poland and Brazil cases we could almost double that, in these high-profile cases, yeah.

AYRES: That’s really something. Joe, comments?

DIELEMAN: Well, so Simon just mentioned there’s essentially two different ways to look at this. The first analysis looked at distinct changes, whereas the second uses statistics that look more at gradual changes or just changes across time. And I think in that one what you don’t really see are countries that stand out, even though there are. Just the methods don’t necessarily stand out. But you start to be able to see the specific diseases. And so while that first analysis focused more on life expectancy, the second analysis and third really started to look at the causes. And I think that list, for me, is particularly interesting, understanding cardiovascular disease, traffic injuries, and whatnot, as the real places where democracy can have a big impact.

AYRES: And so part of what you’re also arguing is that getting a better understanding of where democracy and the feedback loop makes a difference helps in thinking about where there is a possibility for building strong health systems, right? I mean, that’s part of what we’re talking about here, because you can make interventions on infectious diseases that don’t necessarily require more extensive health systems, correct?

DIELEMAN: Yeah. Yeah. So I mentioned earlier the epidemiological transition. Folks that study health finance speak about a health financing transition. And I think that transition is important to think about how you build and create and maintain, and ultimately finance, a health system. And so the health financing transition says that low-income countries—if you look at countries with a low gross domestic product, they tend to generally rely on out-of-pocket spending and development assistance. So that’s out-of-pocket spending at the point of care, which generally doesn’t lend itself to building a robust health system.

And as countries develop economically, and if you look across as you move from low, lower-middle income countries to now upper-middle and high-income countries, you see that most of the financing isn’t out-of-pocket and development assistance, but then it’s prepaid. Normally through a government, sometimes through private insurance. But the point is that when it’s pre-paid, it lends itself to be able to really build a health system that’s sustainable, that’s reliable across time, and doesn’t deter people from accessing it. Out-of-pocket spending generally lends itself to the patient having to make a decision: Do I have this money now? And if I don’t, I can’t get the care I needed. Whereas pre-paid financing lends itself, again, to a system that can accommodate people in the moment, regardless of that decision. Not needing to ask: Do I have the resources for care now? So I think the financing and where the money comes from is ultimately important in essentially how strong that health system can be.

AYRES: Simon, anything to add?

WIGLEY: No, I think that’s good. I mean, I should point out that we do have this kind of multiple methods approach, which may be coming out from the discussion so far, which is deliberate because we were worried about the weaknesses of each, or the flaws that might be pointed out in some of the methods we’re using. So we deliberately used a mixture in so to leverage the advantage of each method to strengthen our ultimate claim, which is why we have these threshold democracy measures, as well as the measure that we use most of all which is this democratic experience measure.

AYRES: So we’re going to open it up now to questions from members. I am obligated to provide a reminder that this meeting is on the record. Please wait for the microphone, we’ve got some mics throughout the room here, and speak directly into it. And please stand, state your name and affiliation, and please also limit yourself to one question and keep it concise, so we can have as many members as possible get the opportunity to ask a question.

Yes, ma’am.

Q: Elizabeth Anderson with World Justice Project.

And I’m interested in what you see in the data in terms of the relationship between democratic backsliding—we talked about the transition to democracy, but now we see it across the globe, democratic plateauing and backsliding, and how does that then shape health outcomes?

WIGLEY: So we didn’t look specifically at the democratic erosion, because that’s a relatively recent phenomenon. I know that in some countries that’s not quite true. So the countries I mentioned earlier that may be eroding or backsliding as recently as 2014, even later. So we didn’t have the option to sort of look at this really interesting case. We looked at the reverse, in a sense, the transition to, and sustained transition to, rather than the degradation of democracy. But this is something which obviously is very important. And we can only infer this by free and fair elections is an important component, as we find, a country that degrades in terms of free and fair elections is not going to perform as well in terms of adult health and noncommunicable diseases, yeah.

AYRES: Sir, you had a question right here.

Q: Thank you. Karl Hofmann with PSI, Population Services International.

I want to congratulate the authors. I look forward to reading this, because it’s a fantastic—I can tell, I fantastic piece of work.

I wonder if the research included looking at something that is not a disease, but which is nonetheless an important driver of morbidity and mortality, and that’s fertility—the relationship between democracy and fertility.

AYRES: Yes, no? (Laughter.)

DIELEMAN: So I think the easy answer, and this is what I would have added to the first question, is we didn’t focus on that. There may be research on that that Simon or the other panelists can highlight. I’m not an expert on that, sorry.

AYRES: That’s a no for Simon?

WIGLEY: Fertility I find tricky. It seems to give competing results. And I find mortality as an ultimate indicator is more clear-cut. But, obviously, that’s not a very interesting reply. (Laughter.) I apologize for that.

AYRES: We’ll get a question back here.

Q: Chris Collins with Friends of the Global Fight Against AIDS, TB and Malaria.

Thank you for this very interesting research. And I’m wondering if there’s anything you could observe about possible effects from the other direction. In other words, over time do societies, as they get healthier, does that have impacts on governments?

WIGLEY: So obviously we didn’t look at this directly, but one could argue that that the development of human capital of all kinds—including health and education—may have a positive impact on not just the economy, but also institutions. So obviously there’s a lot of theory that we would have to work through to position this. So there’s kind of, you know, the modernizing institutions for health. The focus before might have been more on straightforward education, but health—and then now health in terms of the child, for example. But perhaps one implication of our study is the significance of adult health and noncommunicable disease on this issue. The long-run impact on democratic institutions is a very interesting question, I think.

DIELEMAN: Yeah. All I would add to that is more emphasis on the concept of human capital. In particular, on the health side of the contributions to human capital, the focus on childhood diseases, things that interrupt the period of education. And then also adult diseases that prevent laborers and workforce participants. So, again, connecting back to health and economic development is something that certainly researchers continue to point towards, and the science there continues to evolve, so.

AYRES: I saw a question over here.

Q: Hi. I’m Alicia Phillips Mandaville. I’m from IREX.

And this is just brilliant. I used to run the scorecard country criteria system from MCC, Millennium Challenge Corporation, and it used both democracy and health data. So I appreciate the gulf between those datasets is not just wide, but sometimes full of sharks. (Laughs.) So I really enjoy this.

I’m curious about whether you see in the results anything similar about the lag time between the moment of transition and the occurrence of the improvement in health outcomes, because that would be an interesting way to look at kind of hypothesizing around what causes the transition to—or, the outcomes to change.

WIGLEY: Well, the analysis we looked at, the so-called synthetic control method, which means, as I implied before, that you were comparing a country with itself if it hadn’t transitioned, if it had stayed autocratic. What we find, perhaps a little surprising for me at least, is that the effect is almost immediate. So it jumps to, like 2 percent after five years. But it’s not a delayed response. This is immediately shifting towards that 2 percent. And, you know, this is not a clean break. This is—the intervention is not like a medical intervention. This is usually transitions over a slow—over time phenomena. So this was a little surprising for me, at least, that actually the effect was, based on our measurement of democratic transition, was immediate.

AYRES: Yes.

Q: Thank you. Shanthi Kalathil from the National Endowment for Democracy.

I’m really interested in the element of elections versus other components of democracy. And I was wondering if you could speak to—I think the moderator mentioned that media freedom and freedom of expression was not statistically significant, but elections seemed to be the decisive feature of democracy that matter. And I was wondering, given that you pointed to China’s elements of media control and you specifically highlighted the expression elements, why is it that elections in particular are so significant?

WIGLEY: So it’s the share of the population that are potentially included is one part of this. Compared with an autocracy where the support base can be very narrow. It could be the military, it could be a business elite, it could be an ethnic group. So in theory, there will be a larger slice of the population that are sending these vote signals that there is health problems. I realized there’s criticisms of this, but the claim seems to have, for me, a lot of intuitive force. But also, we’re assuming that media freedom is part of the package. It’s not just regular elections, but there’s also criticism through the media or civil society is included in the measurement. So there’s diverse sources of feedback coming from civil society and media, and voters directly.

It is the case that free and fair elections seems to have an outsized impact. But I think that’s to do with the signaling—it can be connected back to the signaling from the voters. So if half the voters or the opposition is effectively not able to win the elections, their signals are effectively eliminated in this case. I should say that free expression was the second-most significant, but by a large shot it was more to do with free and fair elections. Yeah.

DIELEMAN: And, real quick, to just simply add to that, is it’s important to remember—and maybe this didn’t come out in the presentations—that the best—the largest impact was still the composite measure. It was all the pieces put together that lead to the largest health change. Whereas then we took steps to really try to dive into it and said, well, what could be the second-largest impact, when you start pulling out different pieces. And that’s when we came to the conclusions that we’re discussing now. But still, the most important, largest impact was the package of all the components.

AYRES: Sir, question in the back.

Q: Thank you. Carlos Williams, Senator Markey’s office.

We look at now, we have still a lot of low-middle income countries, almost 123 countries, that are democratic in government. And a lot of them are still low-middle income. Is there any association with once you transition to democracy, that opens the doorway in access to foreign assistance at a greater level, therefore increasing health outcomes?

DIELEMAN: Yeah. So there has not been a lot of direct connection between democracy and the receipt of development assistance. We’ve seen in the past that there are characteristics that drive the receipt of development assistance, but democracy or other governance indicators generally haven’t shown a statistically significant effect. And I think that’s one of the takeaways from the paper, at least the things we tried to question, is moving forward in a world where development assistance for health, I mentioned, grew dramatically at the beginning of the millennium is now relatively flat. The year over year growth is pretty small, if even positive. And asking the question: Where should that limited set of dollars be provided, and what can it be a catalyst for?

AYRES: Yes.

Q: I’m Lucy Mize with the Asia Bureau at USAID.

I’m curious if you disaggregated some of the national-level indicators, because if you look at a country like Indonesia, their health equity issues vary widely between Papua and Java, and it’s arguably a democracy. So how did you address that in some of the bigger countries where there are such gross inequities?

DIELEMAN: Great question. So two thoughts. One is, in this study we simply didn’t. We used national-level dataset, and that’s what we focused on. But what you’re drawing out, my second point, is that as far as the health statistics go—and maybe Simon can speak to democracy and local democracy versus national—but as far as the national statistics go, there’s a huge push to measure these things at an increasingly granular level. And so you used Indonesia and India as examples. Both of those are countries that are increasingly we can measure health outcomes by state. China as well, Brazil, the United States, U.K. These are countries that are large enough or have good enough data that we can take the set of outputs that we measure and measure at a quite granular level.

Another project that we have isn’t for all causes of illness, but is for many, and actually measures health outcomes at the five-kilometer by five-kilometer grid throughout sub-Saharan Africa. And that really tends to focus more on infectious diseases, with the idea that you really need to know precisely where an infectious is taking place, where is the vulnerable population and how is it spreading? And so for all of sub-Saharan Africa, we can measure mortality due to infectious diseases and under-five mortality as well at a really granular level, to highlight some of the geographic inequality that we see.

AYRES: I saw two more hands. I saw one here.

Q: Thank you very much. Susan Page (sp).

I just wanted to go back to something that was said earlier, I guess, by Thomas. If the development assistance and other health care that has been provided by foreign aid has not really helped the health care systems in the countries to improve for noncommunicable diseases, what accounts for that? And does that mean that although the money that has been spent has helped to eradicate or diminish the number of communicable diseases, why has the health care system itself not been strengthened? Thanks.

AYRES: I have a great book about that, by the way. (Laughter.)

DIELEMAN: Yeah. And simply that—I mean, I think that’s a very complex answer, certainly something Tom can speak more to. And I think it is—in my mind, again, maybe it’s from an economist perspective, but a lot of that comes down to efficiency. As far as some of the treatments that are provided for infectious diseases, to do so in a very vertical environment—where you are simply giving vaccines and coordinating with a relatively small group of people that are also capable of delivering vaccines—there’s efficiency gains. And I think you see that in some of the other communicable diseases that have received so much support from major donors.

The question is, and this is a major conversation—is how can the systems, because there still needs to be a system to develop a vaccine, it’s just not necessarily integrated in the broader health system—how can those systems that have largely been funded by donors be used to help build the local health system that can now respond to a broader set of health conditions? And I think that is a major task, and a major question that is asked, both by recipient nations themselves, but also something I think donors are increasingly grappling with.

AYRES: I saw one more question over—yeah. One over here, and then we’ll switch to the next panel.

Q: Andrew O’Donohue from the Carnegie Endowment. Thanks for this panel.

One key point Professor Wigley mentioned is that the authoritarian regimes of today differ markedly from those of the Cold War period. And I was curious, is there a variation over time within your data, such that today’s authoritarian regimes, in which leaders such Chavez and Erdogan often dramatically expand social spending and health-care spending, are performing better in terms of health-care outcomes than the often-neoliberal military dictatorship at the start of your data in the 1980s?

WIGLEY: So the two pieces of analysis we have, one about the transitions, would include these earlier types of autocracy. But the detail—you know, most of our analysis is from 1990 to 2016. So we’re dealing with this new phenomenon primarily. The closed autocracies, you know, one-party regimes, or theocracies, or monarchies, are still there but they’re becoming a minority in the autocracy spectrum. But because these fine-grained distinctions, we don’t look at regime types, apart from clearly democratic versus clearly autocratic, because obviously this is an extremely interesting point.

But as I implied, we do look at one of the key indicators of this new regime type, which is the free and fair elections. The electoral autocracies or authoritarian—electoral authoritarian regimes, they sustain themselves through this very factor, right, apart from controlling the media, which is obviously important. You know, slanting the elections to ensure reelection is a key factor. And so although we didn’t split this into the sort of nuanced regime types that you’re referring to, it does seem that the one indicator that I’ll come back to is an important factor in terms of adult health, at least, and noncommunicable diseases.

AYRES: With that, please join me in thanking Simon and Joe. (Applause.) We’re going to transition now to panel two. So we’ll just go off the stage this way, I think.

BOLLYKY: Great. Well, thank you for sitting through the transition. And I’ll thank my colleagues for that terrific job on the first panel.

We’re going to shift gears a little bit and talk a little less about the specific study—although if it comes up I’m happy to have some responses in the Q&A or afterwards. But we’re going to talk primarily now, is we’ve put forward in the study a potential motivation to see more support for international democracy and government promotion. So now we’re going to talk to two of the leading experts about what that support looks like now, both for the U.S. government and internationally, what some of the differences are, and maybe hopefully get to some ideas about how the reach and effectiveness of those programs can be increased.

As I mentioned at the outset, we really have two ideal speakers for this topic. To my immediate right is Kate Somvongsiri, who is the deputy assistant administrator for the Bureau of Democracy, Conflict, and Humanitarian Assistance at USAID. To my further right, just physically, not necessarily politically, is Thomas Carothers, who is the senior vice president for studies at the Carnegie Endowment for Peace. And, again, you have their biographies in your materials.

And I’m going to start with Kate.

SOMVONGSIRI: Oh, apologies. I was—I thought you were starting with Tom. But happy to—happy to take the first crack at this.

So I’ll spend a few minutes just at the top talking about what the state of play is for USG broadly, but then focus most of my comments in terms of USAID’s approach. Lay my biases right on the table at the outset. I’m approaching this obviously with USAID with a development practitioner’s perspective. I am a practitioner. I’ve spent most of my time in the field. So we’ll rely on the academics in this room to keep me honest on some of the data points. And then the other—the other bias is obviously this is—democracy promotion is done in many places not necessarily from a development perspective, but the perspective I’m sharing obviously institutionally is that one. And then wrap up with being a little self-critical in terms of what some of the challenges are and what we can do to be better.

So starting out, on the USG landscape broadly, we—USAID is the leading bilateral democracy assistance donor. About 70 percent of USG assistance in this area is implemented by USAID, the rest by the State Department. Where we are today? Really over the past five years in this field there’s been a steady increase or, at least, flat line, not a decrease in terms of just the general amount of assistance. From FY ’15 it was at 1.9 billion (dollars), is our global democracy budget. And this past—this fiscal year ’19, it went up to 2.4 billion (dollars). So I think that shows you kind of just the level of commitment.

I know talking in a room of a lot of health practitioners those numbers may not seem impressive, but for the democracy community—which, you know, twenty-something years ago we were just basically struggling for our survival of even actually being taken seriously as something to do in the development field. This is really, I think, quite heartening for us. And that is really as a result of very strong bipartisan support on the Hill, with this global democracy directive. In terms of how we work with USAID, I think it’s also important to point out structurally it differs from some other places because we’re grounded with this development focus. We have about four hundred democracy experts in the field, in our field missions, including local host country nationals. And then we have about one hundred in Washington. And so the overall democracy strategy is really very knitted into what the overall development trajectory is in that country.

And I’ll just a couple words on that approach. Anyone who—you know, I have AID colleagues in the room—anyone who’s heard anyone from AID speak in the past several months will not be new to the phrase: journey to self-reliance. (Laughter.) So—but for those who may be newer to it, this what we think—maybe I’ve been in the bureaucracy too long, but it does seem like a very—it is a very bold thing for us to actually state that the goal of foreign assistance is to end the need for its existence. So that—and that’s the idea behind the journey to self-reliance, is really defining the path a country need to take to strengthen its commitment as well as capacity to achieve its—and manage its own development goals.

I think for us, why this is important in the democracy conversation is as part of that journey we use a lot of metrics—not quite MCC-level—but we do have a number of metrics that measure capacity commitment to be able to start that conversation with the countries. And woven into those are a series of democracy-related ones, including those by V-dem and many others, measuring open and accountable governance, inclusion, civil society capacity. So we’ve really taken this as a commitment. And the democracy promotion ethos is really woven tightly into the DNA, I think, of the agency at this point. So that’s one thing in terms of our approach.

I think many of what we’ve been talking about today is something that over twenty years our experience has shown that we’re not doing it just because of the inherent value of democracy. Others may do it for that purpose, right? There’s intrinsic moral value. But for us, as a development agency, we have seen great potential to unlock sustainable improvements in people’s lives, higher level of economic innovation, and to moderate social conflict and violence. And because of that, it’s very much the bread and butter of what we do.

The second approach I’ll mention in terms of where we are in this field, in democracy promotion in AID, is we’ve undertaken over the past, oh, five to seven years a very, very deliberate approach towards integrating democratic principles into other sectors, including health. Health, education, economic growth, principles that I think most in this room are familiar with—participation, inclusion, transparency.

And what we found—and, again, it won’t be a surprise to many people—but most of our colleagues who were working in some of these other social sectors have traditionally had a very technical approach. Not a criticism, I think it’s just the nature of the work, technical approach to dealing with the social outcomes they’re trying to get. And what we’ve really done as an agency is really brought a more political economy analysis lens to it and tried to integrate some democracy principles into the world. So I think that’s an important factor in the approach that we’re taking.

I’ll give a couple of concrete examples on this. I think that will lend itself to this discussion. And I will also say, to confess, that, you know, when we first started doing this, we didn’t really get a lot—there were some really kind of thoughtful, forward-looking colleagues who really saw a lot of inherent value in this. But really some of our other colleagues from some sectors would say: You’re only doing this because your democracy budget is small, and our health budget is huge, and so you’re trying to leverage and integrate. And in part, that was true. And I did find that—(laughter)—you know, frankly, smaller budgets cause us to innovate, right? I think we often say, you know, dollars is a sign of our commitment. But, when you have smaller budgets, you are forced to integrate. But I think over time many of our colleagues in other sectors have really seen the benefit of that.

One really small example of a program in Senegal in rural areas, there was—in a rural area where we were working with our governance and peace program, 44 percent of the children in the area did not have birth certificates. It may not seem significant, but that’s what enabled people to access health care, education, other vital services. So by having a program that specifically worked in four specific communes in rural areas to simplify and streamline a birth registration process—and we had pretty dramatic effects. 20 percent increase overall, hundred percent increase in the health centers itself. And that unlocked the potential to access health care.

I think one of the struggles for us, and that is measuring it, right? I think we’ve had difficulty in convincing some of the colleagues initially because they were being held to certain standards of, you know, you need to demonstrate very clearly numbers of—you know, reductions in certain diseases, vaccinations, et cetera. And by doing the broader health systems strengthening, that line was a little bit more attenuated. So how can you really make that convincing argument? This study, I think, should help go a long ways in that.

So those are two things in our approach. And a big part of our approach is also continuing to try to learn from the evidence that’s out there. Many of our speakers previously talked about the varieties of democracy approach, and how much more rigorous it than previous ones. We have a partnership with V-dem now. I’ll highlight examples from two studies that have come out of there that I think will be interesting for the group.

One is a study late last year indicated that greater political empowerment of women is also significantly associated with lower infant mortality rates, higher life expectancy, and higher average years of education. And then another study that just came out in August that countries with higher respect for civil liberties experience higher economic growth rates—in particular, civil liberties like protection of private property. If you ask me any specific questions about those studies I will not be able to tell you answers. (Laughs.) But they are available. And I think it’s really part of this broader trend of the research on this issue, and what it’s contributing to.

And then the last thing I’ll end with in terms of where things are going and the challenges we’re seeing now and what we can do better. I think some of the previous panel started to talk about this. One thing we’re really grappling with, especially within this administration and the work at this time, is the rise of resurgent authoritarianism. We’ve been talking quite a bit about backsliding as a whole as a trend over the past decade. But resurgent authoritarianism and some of what we’ve talked about in the previous panels in terms of using—authoritarians using the language and the cloak and the veneer of a democracy to legitimize their rules.

Nobody—you know, Maduro, Kabila, nobody says: No elections. We’re just going to, you know, stay in power. There’s an adoption, a cooptation of the language of democracy, and then the veneer of that. And then how do we—how do we counter that? I think we haven’t—I think we need to make a lot more progress on that. I think there’s a lot of thinking on that issue. That also coincides with—just as we see the contestation of democratic institutions and processes and norms in the countries where we work, we’re seeing competing models of development as well.

I think, again, not a surprise to folks in this room, but competing model development that prioritizes—that’s more transactional, quick economic gains for a few in exchange for unsustainable debts. I don’t think we have to name any countries. I think everyone is well aware. But this convergence of resurgent authoritarianism while at the same time the shifting standard of development finance, I think, has brought greater urgency for us.

So let me stop there, and then with the exception of just saying one key takeaway from this for us is let’s divert some global health resources to the democracy sector. (Laughter.) And we’re going to be talking about how we can tackle this more in the future.

BOLLYKY: Great. Tom.

CAROTHERS: Thank you. I’ve been asked to talk about the non-U.S. side of the international community that works to support democracy. There’s such an ingrained habit in the United States of talking about the U.S. role in supporting democracy internationally as a kind of uniquely American endeavor. But I’d say it’s an under-developed intellectual and practical muscle in the U.S. foreign policy community to understand and then take action upon the fact that there’s a much wider community of actors engaged in this. If I said to you that there’s an international community of actors working on global health, you’d say, of course. And you’d probably already have a sense of what the dimensions of that are. If I say to you: There’s an international community of actors working on democracy, I suspect your mind might be blanker, and you’d have a harder time profiling.

What that community is—first, it consists of organizations at three different levels. I’ll be very schematic here. The first is there are probably between twelve to fifteen other governments that are serious about this. They range from Canada and Sweden and the Netherlands and Germany, to the U.K., France, Japan in some ways, and others within the OECD countries. But then there’s also some non-Western democracies that are also trying to give international support for democracy—India, Brazil, South Africa, Indonesia are all engaged in different ways in that. So one level is peer governments.

Another level are intergovernmental organizations, the European Union the biggest and most important among them in terms of democracy support, whose democracy support rivals that of the United States in dollar terms or euro terms. But also another of other significant international organizations, ranging from UNDP, to the International Institute for Democracy and Electoral Assistance in Sweden, to a number of regional organizations, and so forth. And then the third level are international nongovernmental organizations. There’s many of these in the United States, like Betsy’s group on World Justice Project. But there’s many international human rights groups, and transparency groups, anticorruption groups, women’s rights groups, and so forth.

So this is a big, complex community that exists out there. The United States is monetary terms, nobody can take the measure of it because it’s impossible to calculate. It’s just the data isn’t really available. But the United States is probably somewhere between 15 and 20 percent of the funding of that community. It has a couple of characteristics as a community. One is that within the community, people think they’re doing very different things from each other and particularly from the United States. If you go to Germany and meet with the director of democracy policy, or Sweden, or almost anywhere within that community, the first thing they’ll tell you is: Well, we’re not like the United States. And they have a caricatured idea of the U.S. role in democracy, which has two features to it.

One is that when we want to promote democracy we start by invading the country. (Laughter.) That’s sort of step one, and then we see what we can do to help bring democracy. And that step two is we then take the U.S. Constitution and translate it into the local language and export our system and send congressmen over to lecture about how well the U.S. Congress works. (Laughter.) That isn’t actually the case. That isn’t characteristic of U.S. democracy support. But unfortunately, that causes people in the international community to not want to sort of think that they’re aligned with it. But there’s also beneath that the simple fact of kind of national accents in democracy support. So Germans put a very strong emphasis on rule of law support. Sweden puts a very strong emphasis on gender relations. The U.K. is doing a lot on anticorruption . Every country has its view of what’s most important.

But if you go beyond those differences and imagined differences, the community is actually doing very similar things in those places. And it’s a myth that there’s really significant differences. You see that at the country level. If you go to the Democratic Republic of Congo and look over the last five years and say: What’s the community of action doing to support democracy in Congo? Very similar things. And at the country level, they’re often joining up in working fairly closely together. It’s at the next level up above the country where they don’t imagine their working together and don’t imagine they’re doing the same kinds of things.

The other characteristic of the community is it lacks leadership. There’s no natural leader of this community. Nobody’s in charge. There’s no preponderance in a sense of governing responsibility and sort of informally within the community. And the United State has, at various times over the last thirty years, sometimes played a leadership role of a certain type, but it hasn’t been very much for a while now, I would say in different ways. And I don’t mean just the last two years, but before that as well. So this is a community with a very foggy sense of itself as a community, both in terms of what it does and how it acts as a community.

Now, next thing I would say is how is the community doing? Well, it’s in a funk. It’s in a fuck because—at two different levels. First, because the major thrusts of work within this community have all, over the last five to ten years, hit some big walls. Let me describe the four big thrusts of this community are, first, work on free and fair elections. And a lot of progress has been made in the sophistication of efforts to support free and fair elections.

But you run up against two walls. One is that you do excellent work to support a free and fair election, but then the results of that work are sort of steamrollered by local forces, and it proves to not have been very significant. The Democratic Republic of Congo’s a good example. The last election there was actually a pretty bad election that turned out pretty badly. But they decided to just go with it, and all of the international efforts were just kind of steamrollered by local political reality. And that happens with some frequency these days.

And the other wall is that governments in many places are extending terms against their own constitutions and overriding the electoral process. So a country like Bolivia, which has made some encouraging developments in democratizing over the last twenty years, is now stuck with a government that seems determined—a president to extent into his fourth term, despite an obvious constitutional prohibition on doing that. And he’s not the only one. So elections have run against some walls.

The second is political parties. A second big thrust of this area, political parties thirty years into the big wave of democracy over the last thirty to forty years proved to be institutions that seem determine not to be able to reform themselves, and to find a new grip on populations. Citizens around the world are very frustrated with their political parties. They have been for a long time. Parties have had a hard time responding. Sitting here in this country, this isn’t a mystery to us. Parties are struggling as organizations to develop or maintain the kind of relations with individuals, with citizens, that they may used to have had at one point, given that life has changed in many ways.

The third area that’s been frustrating is civil society. Big, mushrooming of civil society in the 1990s and early 2000s, but then starting in the mid-2000s many governments, in both non-democratic but as well as some democratic countries, began trying to put limits on what civil society does, trying to co-opt it, trying to compromise it in various ways, create their own, quote, civil societies. And civil society has been under serious siege for the last ten years. And you see this in many different ways.

And then the fourth sort of thrust or the democracy’s main work on democratic governance systems. And here, an unfortunate number of democratizing countries have been stuck with pretty bad governance marked by state capture and systemic corruption. Think of South Africa, a democratizing country with some very encouraging political participation characteristics, but deep state capture, deep corruption, and great citizen frustration as well. And that’s South Africa which is, you know, a sophisticated country and state capacity terms, compared to many other near-democracies.

So you put those together. Elections are difficult, political parties have trouble reforming, getting traction, civil society’s under siege, and a lot of governance is stuck in some pretty bad patterns of state capture, that’s one level of funk. Is that’s our agenda, what are we not doing right, is the feeling within this community. The second level of discouragement is at a—is a broader level, which is what Kate referred to—is the fact that the larger contest of democracy as both a political idea and a practical political system is facing greater competition and a greater sense in the world that it’s not the only model on offer in the world and maybe there are alternatives that work better. And even if they don’t work better, those who proffer them are more assertive or aggressive and successful at putting their ideas forward across borders.

And that’s got different parts of the democracy community worried. And, again, because it’s not a community that functions in an integrated way, there’s no response from this community. There’s a lot of concern, but it’s not—everyone looks around and says sort of to each other: What are we going to do about this? But nobody really knows in a sense how to tackle that. Whether to do it more at the sort of mezzo level of country by country or more at the sort of geopolitical level as a larger sort of effort by democracies. And there’s very little leadership or action in that regard, I would say.

Now, finishing up, I would say that as the community looks forward one of the biggest and most encouraging developments—which, again, Kate referred to and USAID has actually been an excellent actor in this area—has been trying to connect the democracy transition process or the democracy agenda to developmental objectives, given that citizens are very frustrated with what their democratic governments are producing for them in many countries. The idea is let’s connect more directly democratic change to developmental performance and delivery.

The reason why this has been difficult, and only a few places within the community have really made progress—USAID is one of them in this area—is that the developmental side—it isn’t just—I mean, Kate said politely that, you know, it’s because they take a technical approach. There are two deeper reasons why they’re taking that technical approach. One is that there’s still a profound intellectual debate and disagreement over whether or not democratic governance produces better developmental outcomes. And studies like yours—and this is where I connect this, Tom—are of fundamental importance in that debate, because the democracy community has not won that debate.

And, you know, not long ago I was—a year or two ago I was at the German Development Bank in Germany. And I gave a talk. And, you know, the crowd was outraged. How did I know that democracy would be good for development? You know, just prove it. You know, we have no commitment—we are a development bank. We have no reason to be interested in politics. You know, prove to us that your model can produce better growth rates. And because they focus a lot on growth, not just the German Development Bank but much of the development community, it’s a harder debate if you tried to write your paper or book on just growth in democracy. Although, there is a debate over that. But it’s important when you get to the broader debate, not just growth but development in a more general sense. You can win parts of that argument in important ways.

And then the second reason why the development community is held back is a fear of, quote, “being political,” and therefore not being able to work in sensitive political environments. We wanted to work in the last 10 years in Ethiopia. And we were afraid if we raise our voice about human rights we would be kicked out of doing this great work on child mortality in Ethiopia. And so a desire to stay out of politics. And that’s where the technocratic impulse sort of comes in as well.

And so both these areas of debate are really the cutting edge of where we are in terms of integrating this agenda. And that’s why work like you do is so important, and why leadership that USAID and a few others—DFID and a couple of other organizations, and the Australians have also made some progress in this regard and some other—of trying to integrate is really vitally important. And so it isn’t just about budgets, although budgets are always important at some level. And what’s great about AID is they’ve had significant capacity and technical thinking about democracy, as well as other domains, and the ability to talk within the agency about bringing these things together. And they’re starting to do so. But it’s amazing that thirty years, forty years into this field that we’re only starting to get serious about this conversation of connecting these two areas. But better later than never.

BOLLYKY: Great. So those were two terrific presentations. Thank you for that. A couple of things resonated with me with the remarks, particularly that you just made. I will tell you, that getting this paper published, required confronting two challenges one faces in the global health community—or, three, really—and these challenges are the same as the ones you’ve identified. The first challenge is that most people that we encountered through this research process on the global health side really do view democracy promotion as being focused on invading other nations and imposing democracy afterwards. So there are some serious misconceptions around that. Second, there is a also a general perception that perhaps democracy as a model is not offering very much to other nations right now, given some of the political upheavals happening in many democracies. And the last challenge is there’s a complete lack of confidence that democracies would produce better on health.

So with that said, I want to push back a little bit on the issue of elections. And that’s really for two reasons. One, obviously, it’s one of the results in our paper that stood out the most. Second, I want to be clear about what we found, because this issue came up in the first panel. We didn’t measure just free and fair elections on its own. We took one piece out at a time from the larger democracy measure. So the effects of the various components of democracy are synergistic. You really do need all the elements of democracy. We just measured what would happen if you took one component of that broader measure of democracy at a time. And when you take out free and fair elections, you see that overall health effects of democracy disappears. So, to be clear, it’s not that free and fair elections alone would deliver these health benefits. It’s that taking out free and fair elections eliminate those benefits.

Which gets us to my question. Several people have commented on the global trend toward authoritarianism is driven by election-fixing. So let’s stipulate, just for the sake of argument, not to associate you two with our paper findings, that to produce these health gains it really is free and fair elections have to be present. Kate, you spoke very powerfully on what can be done to try to increase participation and other good governance elements of health programs. And Tom, you spoke a lot about who the community that support democracy promotion. And it’s the same donors -- U.S. government, the U.K., Japan, and Sweden—that are also all big global health donors too. If you accept this paper’s results that the sustainability of our global health gains, on which global donors have spent $40 billion per year to achieve, those gains depend on more free and fair elections – how do we achieve that?

SOMVONGSIRI: Yeah. So I’ll start with three comments related to that. One, which I think will be more difficult to do than perhaps the others is—and Tom alluded to this, I think—to the extent that we can clarify the misconceptions around our electoral assistance. You will not be able to depoliticize election assistance. That’s just inherently incompatible. But there is a misconception that—and, again, I’m going to talk from a USG perspective—that USG electoral assistance is about trying to achieve a certain outcome. We want it—this is the horse we’re backing. Our assistance is going to get so-and-so in power. And what we strive to do over and over, and sometimes less successfully than others, is to clarify what we’re trying to do is an electoral process. That is transparent, that’s inclusive, et cetera, and that it is not about choosing winners and losers. So we need to do a better job, and I don’t know how successful that’s going to be. But we do need to keep emphasizing that.

The second part of it I think we need to also expand—I think most people who work in this community know it well. And I think this is probably tied to your dataset. Elections, we’re not talking about an electoral event. We’re talking about the enabling environment in the lead-up to elections, which is years and years in advance. And this may be—I’m completely speculating, so this is dangerous territory—but in terms of your data, that may be why the election piece is so critical and significant, because you need all these other pieces—like the media environment, freedom of expression, et cetera—to actually achieve that free and fair election. So it may really be kind of so looking at that more broadly.

And then the third thing I would mention in this realm is I don’t know what the V-dem data actually looks like on this, but I think it’s very important to have a conversation about not only national-level elections but subnational elections, which is a big part of our work that doesn’t get as much attention. But my bet, if I were to bet on the study, would be to look at if you’re going to look at linking the causality with the elections, most national elections are not decided on whether—how much the government is spending on health, right? But I would—I would imagine that it’s really more of the local elections, which is much closer to the ground, which is much closer to what the different citizens and response are that some of those linkages may be clearer. That’s what we found in local governance work broadly, and which is why we invest so much in this. So those would be three ideas.

CAROTHERS: Yeah. As the international community focused on elections increasingly over the last thirty years, it’s kind of—excuse the crudeness of the analogy—but it’s sort of a tube of toothpaste. And the international community grabbed one end of it and all the toothpaste went to the other end. What do I mean by that? As we’ve focused on helping countries hold better balloting and be more capable and about the tail-end of the process, the bad things move to the other end. They are exclusion of candidates, intimidation of candidates, use of state resources, and then dominance of the media space, and disinformation, manipulation.

Going after a government that’s doing those things is much harder than meeting with its electoral commission and saying: You’re going to need this many ballot boxes on election day because we’ve helped you figure out the number of people who vote per hour and so forth. Both it’s more political to get at those things, and the standards are less clear. What are the standards about, you know, how much money does each side have to spend? You know, you get into debates about the political financing and what’s the right system, what’s the standard of media access? Well, the U.S. has debates. Other countries don’t have debates. How do you—you know. And so you’ve lost the certainty of standards that you have at the other end of the tube of toothpaste.

And so most of the bad things in elections are happening up there. And then the second thing that’s happened is governments then, if they can’t control it and end up losing, then they go ahead and just trump the process, like happened in Congo, and just sort of steamroller it. And that takes the international community to say: We really aren’t kidding. We’re not going to deal with this new government in at least some significant ways, or reduce our commitments to it, as a result of that action. And that’s been, you know, there isn’t—because there isn’t a very joined-up international community, governments know they can do that kind of thing. And DRC knows it’s going to probably keep getting the international assistance it had before, because deep down they feel like, you know, the client that owes so much money to the bank, that the bank’s going to keep lending them more money.

And so the international community needs to be more political in elections work in this way, and more pointed, and then more serious in a joined-up way at the other end of the process. That’s what would really need to happen to make it better.

BOLLYKY: Great. I’m going to ask one more, two-part question, then I’m going to turn it over to the audience. So start thinking of your burning questions now, and I will call you in the order that I see your hands.

So two things come to mind. One is the Kenya circumstance, where you had an election that was originally certified and then found afterwards that perhaps it had been manipulated, but in a more sophisticated way than elections had been manipulated there in the past. And I’m curious about our data and the tools we have to address that situation. Have we gotten better as an international community at identifying the more sophisticated ways of election rigging> The second question is based on—and this is a good opportunity for me to plug—a book that Tom wrote in 2013 or ’14. In fact, it’s one of my favorite books on international development.

One of the suggestions that Tom’s book makes, which we cite and reinforce in the paper, is the notion that selective funding, the MCC model, may be the way forward on democracy promotion. You often can’t withdraw global health funding. For instance, you can’t pull PEPFAR funding when people are on treatment and we’re saving children’s lives. So it’s only when donors are shifting funding to these new health causes that we could identify and favor those nations that are more promising based on their democratic status. Are we seeing that trend at all in global health? Is there a good example of where that’s starting to occur?

SOMVONGSIRI: I’ll take the first part of your question, in terms of the Kenya example. Absolutely. And I think this is what we’re referring to in terms of the most sophisticated way—like, nobody’s going around stuffing ballots. Well, some people are. But that’s not really—the very crude way, right? It’s happening with the Kenya example—

CAROTHERS: North Carolina. (Laughter.)

SOMVONGSIRI: You could say that. Internationally in our development experience—(laughter)—we’ve seen more sophisticated ways of stealing an election. In the Kenya example, right, the role of the court, and is that going to set a precedent elsewhere? So I do think the answer—the short answer is, in terms of our response, it’s half and half. Yes, we’ve gotten better at identifying these things. We see observers who are faux observers, and, you know, what are they actually documenting? So we’ve seen more and more about that. So recognition is the first part of it. In terms of our tools to actually address that, I think we’re still a long ways off in terms of actually addressing those effectively. And that’s what we really need help on in this community.

BOLLYKY: Great. Tom.

CAROTHERS: Regarding your second question, I’d be curious—Alicia’s here. She asked a question before. She had the long experience at the MCC of—I think when the MCC was created we hoped that that approach, kind of positive conditionality, if you will, giving more for doing better, would spread in the international aid community, because it can become exponentially more effective if it’s not just the United States with a relatively modest MCC budget doing it, but a joined-up approach, in which many do. And that’s part of the problem of the development community not working effectively often as the community, by not being able to do that. I just—both on the experience of positive conditionality on whether and why it has become an approach that others have joined. And she would be better to answer than me. So maybe she can be handed the microphone at some point.

BOLLYKY: Great. Well, I will let her volunteer for that role. But I am going to turn it over to the audience for questions. So either on MCC in particular or to broader questions on democracy and health, if you’ll raise your hands I will call on you.

Please.

Q: Hi. Elizabeth Anderson, World Justice Project.

I was intrigued that you both kind of doubled down on integration. And I say that as somebody who drank that Kool-Aid and believed—I believed that it’s important, and yet have hit my head against that wall and considered it really challenging and seen how difficult it’s been. And hearing about this paper I thought: Oh, my gosh, beyond integration. We can be liberated. Now we have evidence that it’s just elections. We don’t actually have to integrate anymore to see results here. So I’d been interested in your reaction to that thought.

SOMVONGSIRI: Absolutely not. You’re not liberated. (Laughter.) No, I think—well, I think one thing I want to be—make sure I caveat the statement around is this is obviously very specific to certain specific health outcomes, and the integration agenda is obviously more broadly across a range of social sectors. So I would want to be careful about how I compare that. The other thing I think in terms of integration is because the breadth of the work that we do in development is so large, and I think we—I think you’re right in terms of we do have a long ways to go in terms of demonstrating the tangible impact of an integration agenda. But I think there’s a lot of—this is the instrumental versus kind of intrinsic value of democracy argument that many of you are familiar with, right? I think there are many of us who would say there’s inherent value in more participatory approaches towards whether it’s education, or health, et cetera, anyway. So regardless, obviously we’d want it to lead to strong development outcomes as well. But regardless of that, I think we’d say there’s some value in that approach, regardless.

CAROTHERS: This gives me a chance to mention the title of my book that you mentioned in case you’re interested. It’s called Development Aid Confronts Politics. And the subtitle is The Almost Revolution, about whether or not this attempt at integration will ever really revolutionize assistance and get over this gap between the political and the socioeconomic that’s bedeviled the assistance community. And I wrote the book at a time I was moderately pessimistic. And I think I’m still moderately pessimistic, because the forces of—that want to just keep doing development assistance in certain ways that are sort of nonpolitical in bad ways are powerful.

But on the other hand, I’d say that it’s—big changes in development work don’t happen in five to seven years. They take a long time to really change. And so we shouldn’t give up, because it’s the right—it’s one of the right fights to be fighting. It’s important to do this. And I think it’s just more that this study, if he’d found this across ten different core socioeconomic sectors, maybe we could just have the big party and say it’s over, you know, we’ve won. But we haven’t. You know, it’s one good study and such, but there’s the forces that both think that, you know, democratic governments are just freewheeling governments that spend money badly. And you can’t trust them. You need the firm hand.

And then the larger sort of authoritarian development model is so strong these days that we’re a long way from winning the battle at that level. So we both have to fight it at that level, but then also show that a good health program that thinks of why, in this region, is health spending lower than here? Oh, because of the ethnic composition of this region is that, and the president’s family is disfavoring this ethnicity because of the political compact they have. And so therefore a smart health program has to also deal with the political settlement in the country. That’s the level we’ve got to get serious about and start really integrating political thinking into health education, agricultural land, housing, et cetera.

BOLLYKY: Great. Please.

Q: Again, thank you very much. I wanted—

BOLLYKY: Can you just remind them of your name here?

Q: Susan Page (sp). Sorry.

So I want to go back a little bit to the way that you, Tom, described sort of the community. And in talking about the global community, whether it’s linked to democracy and health, why hasn’t there been a broader response to these undemocratic elections? So whether from the U.S. government, other countries, or the United Nations—which has also just recently revamped its development assistance, its development model, along with the peace and security, et cetera, et cetera. But nobody really came out against the elections in DRC.

And in fact, the secretary-general basically said: These were—you know, congratulations to the people in the population of DRC for their good elections. I mean, this is from supposedly the—you know, the head of the global community. How can we get back on track? And I understand—you know, as a lawyer, I understand the desire to sometimes be technical as to avoid getting involved in the politics, which is impossible, as you’ve said, in elections and other political issues. But how can we gain back some sort of moral authority more collectively?

CAROTHERS: What time is it? (Laughter.) That’s a big one. Look, there are different reasons why what happened happened in DRC in terms of international response. The best argument is a functional one of there’s just really no choice. It would just unsettle things so much politically if the international community really just said, no, we don’t want to deal with this government because we don’t believe that it emerged victorious out of that election. And there’s the fear that would destabilize a country that’s been through so much instability in damaging ways.

But there’s another level up, which is still the sovereignty argument of kind of we can tell them what we think went wrong, but it’s kind of up to them as a country—which is true at one level. But the sovereignty, given that we are in an age when sovereignty is kind of having a comeback a bit, we’ve sort of lost that fight a bit, even on human rights, of saying they’re universal rights that every country really has to respect if it wants to be a good-standing member of the international community and so forth.

So I think it’s both in some cases, sometimes just a calculation country by country that ends up creating enough bad examples that others can point to. But then we also have been slipping backwards in terms of winning the larger argument that, you know, actually there are standards that are really—are meaningful, both in instrumental and in intrinsic terms. And unfortunately, that has not done well in the last ten or fifteen years.

BOLLYKY: Great. So we have two minutes and one last question.

Q: Lucy Mize with the Asia Bureau, USAID.

I would just like to make a point that if the global health community is going to benefit from gains in democracy, the global health community needs to stop thinking about health as a disease process. And our funding streams still come by MCH, HIV, Malaria aid, but the fact of the matter is political—development is political. And global health gains come from outside the health system, along with all the improvements in the health system. That is a fundamental flaw in our thinking. And I would just caveat that this is my personal opinion. (Laughter.)

BOLLYKY: All right. Great. So we’re out of time. The one last thing I want to say about this, I’m hopeful this is a conversation between the global health and democracy sectors that can continue. We talked a bit before about what the burden is of infectious disease and that it’s ten million deaths a year. Cardiovascular diseases alone killed fourteen million people globally. So these noncommunicable diseases are a big issue moving forward. And I’m hopeful that this paper, but more importantly these events can start to lead to more dialogue between these communities on how we start to view global health also as a problem of governance and health systems, as opposed to just diseases and discrete interventions such as vaccines. Thank you very much. (Applause.)

(END)

This is a corrected transcript.