Interesting academic articles for March 2020

Here are some of the things I’ve found interesting in the last month!  Happily, none of it’s on coronavirus, and probably won’t be for a while.  The types of large, experimental studies or deeply historically grounded studies which interest me don’t have very rapid turnaround times.

Rachel Sweet.  2020.  “Bureaucrats at war: The resilient state in the Congo.”  African Affairs.  

Rebels often portray themselves as state-like to legitimize their rule, yet little is known about their on-the-ground relations with the administrators of state power—official bureaucrats. Drawing on internal armed group records from the Democratic Republic of Congo, this article argues that rebels’ state-like image is more than a simple veneer: Bureaucrats actively sustain state institutions and recruit rebel support during war. It develops a theory of the sources of leverage that bureaucrats use to negotiate with rebels. These interactions entail dual struggles to sustain the structures and symbols of state power and to shape the distribution of control over these institutions during war. On first front, bureaucrats can use their official status to market the symbols of state legitimacy—official certificates, codes, and paperwork—to rebels. On a second, to recruit protection for administrative posts. Pre-existing routines of noncompliance, like parallel taxes and sabotaged information, can use bureaucratic discretion and opacity to limit rebels’ takeover of state structures. This view from the ground demonstrates the real-time continuity of bureaucratic practice through daily paperwork and exchange during war. It contributes to research on rebel governance by illustrating new competitions for wartime statehood and illustrates the empirical practices of states seen as ‘juridical’ or weak.

Jeremy Bowles, Horacio Larreguy, and Shelley Liu.  2020.  “How Weakly Institutionalized Parties Monitor Brokers in Developing Democracies: Evidence from Postconflict Liberia.”  American Journal of Political Science

Political parties in sub‐Saharan Africa’s developing democracies are often considered to lack sufficiently sophisticated machines to monitor and incentivize their political brokers. We challenge this view by arguing that the decentralized pyramidal structure of their machines allows them to engage in broker monitoring and incentivizing to mobilize voters, which ultimately improves their electoral performance. This capacity is concentrated (a) among incumbent parties with greater access to resources and (b) where the scope for turnout buying is higher due to the higher costs of voting. Using postwar Liberia to test our argument, we combine rich administrative data with exogenous variation in parties’ ability to monitor their brokers. We show that brokers mobilize voters en masse to signal effort, that increased monitoring ability improves the incumbent party’s electoral performance, and that this is particularly so in precincts in which voters must travel farther to vote and thus turnout buying opportunities are greater.

Darin ChristensenOeindrila DubeJohannes Haushofer, Bilal Siddiqi and Maarten Voors.  2020.  “Building Resilient Health Systems: Experimental Evidence from Sierra Leone and the 2014 Ebola Outbreak.”  Center for Global Development working paper no. 526.

Developing countries are characterized by high rates of mortality and morbidity. A potential contributing factor is the low utilization of health systems, stemming from the low perceived quality of care delivered by health personnel. This factor may be especially critical during crises, when individuals choose whether to cooperate with response efforts and frontline health personnel. We experimentally examine efforts aimed at improving health worker performance in the context of the 2014–15 West African Ebola crisis. Roughly two years before the outbreak in Sierra Leone, we randomly assigned two social accountability interventions to government-run health clinics—one focused on community monitoring and the other gave status awards to clinic staff. We find that over the medium run, prior to the Ebola crisis, both interventions led to improvements in utilization of clinics and patient satisfaction. In addition, child health outcomes improved substantially in the catchment areas of community monitoring clinics. During the crisis, the interventions also led to higher reported Ebola cases, as well as lower mortality from Ebola—particularly in areas with community monitoring clinics. We explore three potential mechanisms: the interventions (1) increased the likelihood that patients reported Ebola symptoms and sought care; (2) unintentionally increased Ebola incidence; or (3) improved surveillance efforts. We find evidence consistent with the first: by improving the perceived quality of care provided by clinics prior to the outbreak, the interventions likely encouraged patients to report and receive treatment. Our results suggest that social accountability interventions not only have the power to improve health systems during normal times, but can additionally make health systems resilient to crises that may emerge over the longer run.

Wei Chang, Lucía Díaz-Martin, Akshara Gopalan, Eleonora Guarnieri, Seema Jayachandran, and Claire Walsh.  2020.  “What works to enhance women’s agency: Cross-cutting lessons from experimental and quasi-experimental studies.”  J-PAL working paper.

Women’s agency continues to be limited in many contexts around the world. Much of the existing evidence synthesis focuses on one outcome or intervention type, bracketing the complex, overlapping manner in which agency takes shape. This review adopts a cross-cutting approach to analyzing evidence across different domains and outcomes of women’s agency and focuses on understanding the mechanisms that explain intervention impacts. Drawing from quantitative evidence from 160 randomized controlled trials and quasi-experiments in low- and middle-income countries, we summarize what we know about supporting women’s agency along with what needs additional research.

Tom Lavers and Sam Hickey.  2020.  “Alternative routes to the institutionalisation of social transfers in sub-Saharan Africa: Political survival strategies and transnational policy coalitions.”  Effective States in International Development working paper no. 138.

The new phase of social protection expansion in the Global South remains poorly understood. Current interpretations use problematic evidence and analysis to emphasise the influence of elections and donor pressure on the spread of social transfers in sub-Saharan Africa. We seek a more nuanced explanation, testing an alternative theoretical and methodological framework that traces the actual process through which countries have not just adopted but institutionalised social transfers. Two main pathways emerge: one involves less electorally competitive countries, where the primary motivation is elite perceptions of vulnerability in the face of distributional crises, augmented by ideas and resources from transnational policy coalitions. The other entails a primary role for transnational policy coalitions in adoption, before competitive elections and the need for visible distribution drive institutionalisation. Consequently, the latest phase of social transfer development results from the interplay of political survival strategies and transnational policy coalitions.

Karthik Muralidharan, Paul Niehaus, and Sandip Sukhtankar.  2020.  “Identity Verification Standards in Welfare Programs: Experimental Evidence from India.”  NBER working paper no. 26774. 

How should recipients of publicly-provided goods and services prove their identity in order to access these benefits? The core design challenge is managing the tradeoff between Type-II errors of inclusion (including corruption) against Type-I errors of exclusion whereby legitimate beneficiaries are denied benefits. We use a large-scale experiment randomized across 15 million beneficiaries to evaluate the effects of more stringent ID requirements based on biometric authentication on the delivery of India’s largest social protection program (subsidized food) in the state of Jharkhand. By itself, requiring biometric authentication to transact did not reduce leakage, slightly increased transaction costs for the average beneficiary, and reduced benefits received by the subset of beneficiaries who had not previously registered an ID by 10%. Subsequent reforms that made use of authenticated transaction data to determine allocations to the program coincided with large reductions in leakage, but also significant reductions in benefits received. Our results highlight that attempts to reduce corruption in welfare programs can also generate non-trivial costs in terms of exclusion and inconvenience to genuine beneficiaries.

Matteo Alpino, and Eivind Moe Hammersmark.  2020.  “The Role of Historical Christian Missions in the Location of World Bank Aid in Africa.” World Bank Policy Research working paper no. WPS 9146.  

This article documents a positive and sizable correlation between the location of historical Christian missions and the allocation of present-day World Bank aid at the grid-cell level in Africa. The correlation is robust to an extensive set of geographical and historical control variables that predict settlement of missions. The study finds no correlation with aid effectiveness, as measured by project ratings and survey-based development indicators. Mission areas display a different political aid cycle than other areas, whereby new projects are less likely to arrive in years with new presidents. Hence, political connections between mission areas and central governments could be one likely explanation for the correlation between missions and aid.

Using markets and donor support to mitigate the economic effects of coronavirus in African countries

ETA: a shortened version of this post has been published at African Arguments.

One truism about the coronavirus pandemic is that it is amplifying pre-existing vulnerabilities and inequalities.  This comes through particularly clearly in discussions about how African countries should respond to the pandemic. Rich countries have a range of options for containing the virus and mitigating its economic impacts but these solutions often require money or public health capacity which African countries lacked even before the pandemic.

This doesn’t mean that there is nothing to be done, however.  There’s been an outpouring of research on approaches to pandemic control and economic relief which are tailored to African contexts.  In particular, it may be possible to mitigate the economic impacts of the crisis through a mixture of targeted investments in public markets and public transport, and donor-funded cash transfer programs.  

How are rich countries responding to the pandemic?

To generalize a bit, rich countries have considered four different responses to the pandemic.  Most of them are clustering around approaches 3 or 4 at this point, which contain the virus and minimize economic disruption, but require lots of spending and strong public health capacity.

  1. Uncontrolled spread: Letting the pandemic spread mostly unchecked in order to rapidly build immunity in the population. This leads to many avoidable deaths, completely overwhelms the healthcare system, and causes a major economic crash, so that’s clearly a bad idea.  The UK briefly considered this before changing their tune.
  2. Lockdown without income replacement: putting the country on lockdown in order to avoid the spread of the virus, without compensating people for lost income.  This leads to higher rates of infection because many people must violate lockdown to continue working and feeding their families, and leads to an economic contraction as many others lose their jobs.  The US is only replacing a fraction of people’s lost incomes with a one-time $1200 check, so it’s a good example of this approach.
  3. Lockdown with income replacement: self-explanatory.  There are a variety of ways to replace lost incomes, from paying companies to keep their workers on board (like Denmark) to offering direct cash transfers to people who’ve lost their jobs (like Canada).  This approach slows the spread of the virus while preserving people’s ability to access the goods and services they need to stay healthy.  However, it’s also extremely expensive.
  4. Testing and containment: Testing enormous numbers of people in order to contain infections before they spread, and letting people who are not infected continue to work and travel as usual.  This is the Taiwan-Singapore approach, and it’s been successful at constraining the spread of the virus while minimizing economic disruption.  However, it relies on the existence of strong public health systems and access to hundreds of thousands of coronavirus tests.

What are the options for African countries?

Which of these approaches might be useful for African countries? Approach 1, uncontrolled spread, is inherently a bad idea, and many governments are already taking steps to avoid this.  South Africa and Uganda are on national lockdown already, Kenya is encouraging cashless transactions, Ghana is shutting down markets to fumigate them, and so forth.  We can also rule out approach 4, testing and containment, as most countries don’t have the extensive public health infrastructure needed to do this.

This leaves us with approaches 2 and 3, lockdowns without or with income replacement.  Right now, many African countries are defaulting to approach 2, as they’re cutting back on economic activity without replacing lost incomes.  This is clearly not sustainable. Many people are subsistence farmers who depend on markets to access inputs like seeds and fertilizer, or live in poor urban neighborhoods without regular access to water and food even at the best of times. Some governments are already using violence to try to enforce these unpopular lockdowns.  We can’t lose sight of the fact that hunger and violence are also threats to public health.

The best remaining option is approach 3, lockdown with income replacement.  Most African governments can’t afford to massively scale up their welfare systems in a short time.  But with a combination of donor support and targeted interventions to keep markets open while protecting vulnerable people, it may be possible to keep people from going hungry while also lowering coronavirus risk.

Protecting people’s incomes

Let’s start with options to keep people’s incomes stable, as it’s going to be impossible to promote any sort of social distancing if people can’t feed themselves while doing so.  The two options here are promoting a baseline level of regular economic activity, and giving people direct transfers of cash or food.

One option for promoting regular economic activity in a safe way is reconfiguring the physical spaces of retail markets so that people can continue to buy and sell with greater social distance.  This includes expanding the footprint of markets to allow for proper social distancing, installing handwashing facilities, and training retailers in safe product handling practices.  It may also be possible to do staggered lockdowns in various neighborhoods in order to keep markets open while controlling the total number of people entering at any given time.  Public transport remains a point of vulnerability, as many people have no alternative for getting around besides crowded mini-buses or motorcycle taxis, but it may be possible to distribute masks and hand sanitizer at bus stops or via transport unions to lower the risk of transmission.

Even if a baseline amount of economic activity can be safely maintained, many African citizens are still going to lose their livelihoods as demand for things like tourism and agricultural exports drops.  There’s a clear need for direct income replacement from the government. Since most people work in the informal sector, a Denmark-style approach to paying employers to keep their employees on isn’t feasible.  Instead, the best solution is to provide cash transfers directly to individuals.  (If agricultural markets begin breaking down, in-kind transfers of food may also be useful, but these are more difficult to organize than cash transfers, so they shouldn’t be the first step.)

The good news is that almost all African countries have existing social protection programs which offer cash transfers to poor citizens.  These programs do tend to have very limited reach, supporting only small percentages of the poor, but at least the infrastructure exists.  There are also cash transfers run through humanitarian aid organizations and through NGOs like GiveDirectly.  As of March 27, only a few African governments had announced plans to scale up their social protection programs.  However, most countries are early enough in the progression of their pandemics to be able to do this before infections peak.

African governments and aid donors should be immediately focused on scaling up the infrastructure for universal cash transfers.  Many African governments can’t unilaterally afford a huge increase in welfare expenses, and will face falling tax revenues during the period of the pandemic, so this is the ideal time for the major donors to step in and support them — ideally with grants rather than loans.  Time is really of the essence here.  Food prices have already begun rising, and will only continue to do so as imports get delayed or shut down, and domestic food supply is threatened by market failures.  Targeting cash transfers to the poorest also takes time, and in this case many people who were not previously extremely poor are about to lose their incomes, so targeting doesn’t really make sense from an ethical perspective either.

Protecting people’s health

But what about public health in all of this?  If markets and transport remain open, even with social distancing measures in place, there’s clearly still a risk of coronavirus transmission.  Many African countries have only a few dozen ICU beds for millions of citizens, and it’s going to be difficult to scale that up rapidly, given that every single country around the world is trying to procure additional medical equipment on short notice via already-stretched supply chains.  The Africa Centres for Disease Control and Prevention appear to be doing as much as they can to help countries acquire coronavirus tests and protective equipment for healthcare workers, but they can’t compensate for years of low investment in health systems overnight.

A research group at the London School of Hygiene and Tropical Medicine has proposed that public health resources in low income countries should be focused on protecting the most vulnerable populations, rather than trying to stop the spread of the disease generally.  This involves identifying people who are immunosuppressed or over 60 years old, and providing them with a package of services (like a safe place to stay and regular food deliveries) so that they can remain isolated from others during the course of the pandemic.  Younger people can continue to show up to work and keep markets open while they are healthy, and if they become ill, there will be more hospital beds available to them.

This strategy obviously has some limitations.  Coronavirus is posed to spread most rapidly in dense urban neighborhoods, where few people can afford an entire room to themselves and there’s no reasonable prospect of building new shelters.  The prospect of leaving home to go into a state-run quarantine facility is clearly undesirable unless the facility is well-run, which many will not be.  In addition, people with conditions like HIV or TB already face stigma, and may not want to self-identify in order to go into coronavirus quarantine. However, this doesn’t mean that quarantine interventions are doomed to fail.  Research on the 2014 – 2016 Ebola epidemic in West Africa suggests that building community trust is an essential part of developing effective, locally contextualized interventions.  Similar trust-building practices may be necessary to develop successful public health strategies in the age of coronavirus.

 

What’s the risk of coronavirus in Kenya and Uganda?

As the novel coronavirus (COVID-19) has marched through China and Europe, Uganda has avoided any cases to date, and Kenya has only had three cases Like everyone else in the region, I’ve been wondering whether this is likely to continue, and what might happen if the disease did arrive at scale.  I wanted to collect a number of resources that I’ve found useful in thinking about this in a single place.  I am not an epidemiologist and am not making any predictions about the geographic spread of the disease.

Is coronavirus likely to spread to the region?

There are two factors which may play to Kenya and Uganda’s advantage when it comes to preventing the arrival of additional coronavirus cases in the area.  First, the virus has been spreading globally when infected individuals travel between countries.  Both countries have relatively few direct flight links outside of the continent.  Kenya has closed its borders to non-residents as of March 15, and is requiring all travelers to self-quarantine for 14 days.  Uganda is requiring travelers from countries with current coronavirus cases to self-quarantine as well.

World-airline-routemap-2009
Map of global flight paths from Wikipedia.  It’s from 2009, the most recent map available, but the general patterns presumably still hold true today.

Second, in countries currently affected by coronavirus at scale, the disease has also been spreading among communities in contact with an infected person.  However, there’s some tentative evidence that this type of spread is only happening in relatively cool climates. A new paper from Sajadi et al. (2020) notes that all cases of community transmission have been concentrated in a narrow band with temperatures around 10 degrees Celsius / 50 degrees Fahrenheit.  This may be because the coronavirus can’t survive as long outside the body in warmer temperatures.

If this finding holds true, it suggests that if additional infected individuals do travel to Kenya and Uganda, the coronavirus may not spread significantly beyond them.  However, there is still a great deal of uncertainty here, and we shouldn’t trust that warm weather will prevent community transmission.

Are Kenya and Uganda prepared to track the spread of the virus?

There’s reason to be optimistic about public health capacity to track potential coronavirus cases.  Both countries already have infectious disease surveillance infrastructure in place at the international airports due to the recent Ebola epidemic in the nearby DR Congo.  All travelers to Nairobi and Entebbe must report their travel histories, share contact information, and go through a thermal screening at both airports.  Of course, airport screening won’t stop all infected people from entering the country because some may not have symptoms yet, but it’s still useful for surveillance.

In addition, both countries are building on much longer histories of population-level disease surveillance, including those for polio and HIV / AIDS.  As this article from Think Global Health notes,

Because of the robust responses to these diseases, many African countries are starting from a very different baseline than twenty years ago. Although this has not generally included support for ICU-level care that will be required by the sickest people with COVID-19, what these investments have supported are increasingly human resource for health, supply chains, information and surveillance capacities for prevention, detection and long-term response capacity against diverse infectious threats.

Finally, because the onset of the epidemic has been delayed here, both countries have had more time to prepare.  The Africa Centres for Disease Control have been running online trainings for healthcare workers about the coronavirus, and have also developed and distributed coronavirus test kits to most countries across the continent, including Kenya and Uganda.  Both countries have already banned international conferences, and Kenya has also opened a 120-bed isolation center for potential patients. 

What happens if coronavirus does spread within the region?

If community transmission of coronavirus does occur within Kenya and Uganda, one of the main risks is that it may overwhelm healthcare systems with people seeking care.  Taking early preventative measures to slow the spread of the virus makes it more likely that sick people can access care when they need it, as this graph from Our World in Data shows.  Unfortunately, Kenya and Uganda are grappling with weak health systems and poverty, both of which may make it more difficult to contain the virus if it does arrive.

A graph showing that

In both countries, health systems tend to be underfunded.  In Kenya, significant revenue comes from user fees, which discourage poor people from accessing healthcare.  Uganda abolished user fees for healthcare in 2001, but poor people still find it difficult to access care. Kenya is doing fairly well at providing essential medicines, but Ugandan clinics often lack drugs.  The number of intensive care unit (ICU) beds in both countries is low, and only 23% of Ugandan ICUs have ventilators.  On any given day, nearly half of healthcare workers in both countries are absent from their jobs, often because of poor pay and long commutes.  If coronavirus spreads among the population and leaves many people in need of hospitalization, it’s clear that the health systems will struggle to keep up.

Poverty can also make it difficult for people to take other steps to keep themselves safe from coronavirus.  Only 14% of Kenyans have access to soap and water at home, and in Uganda only 8% of families with young children have access to soap and water at home.  Hand sanitizer is available in shops, but not widely used.  Without well-functioning state-run social safety nets, most people also don’t have the luxury of taking time off from work to rest if they are sick.  This both increases the risk that the disease will spread, and makes it more difficult for infected people to recover. Informal insurance within families and religious groups can mitigate this somewhat, but churches have also been vectors for infection in the US and South Korea, meaning that this insurance mechanism could possibly increase the risk of contacting the disease.

Even without cases of coronavirus in the region, the social and economic impacts of the disease are already being felt.  Xenophobic statements about Chinese residents have been reported in Kenya.  And in both countries, an economic slowdown is expected as traders have been cut off from Chinese imports.  The indirect economic effects as people lose their livelihoods may be just as serious as the disease itself.  It’s important for both countries to continue expanding their social safety nets and ensure that healthcare is affordable, even if coronavirus never comes ashore. 

What have previous flu-like pandemics looked like in Africa?

It occurred to me today that I hadn’t heard much about how previous flu-like pandemics have impacted African countries.  Being a geographer at heart, I thought I’d look up some maps.  I’m not an epidemiologist and am not making any predictions about how prior patterns of disease spread might replicate with the novel coronavirus / COVID-19.

I started with Wikipedia’s list of 20th and 21st century flu pandemics, and added seasonal influenza and severe acute respiratory syndrome (SARS), which isn’t the same as the flu.  Let’s work backwards in time, since air travel has become much less expensive and more common since the 1980s, and this really changes the modality by which disease can spread between countries.

The three post-1980 epidemics are coronavirus (2020), H1N1 / swine flu (2009), and SARS (2003).  Despite their weak health systems, African countries have not been highly affected by coronavirus to date.  Transmission of H1N1 was also fairly limited, and transmission of SARS appears to have been almost non-existent.  There’s a lot of debate about why coronavirus hasn’t spread, but two possibilities are the continent’s relatively weak air traffic links to other regions, and improvements in public health surveillance capability since 2000.

Coronavirus / COVID-19 (as of 12 March 2020)

Map showing that the outbreak is concentrated in China, Iran, Italy and the US
Source: Foreign Policy

H1N1 / swine flu (2009)

1280px-H1N1_map_by_confirmed_cases.svgScreen Shot 2020-03-14 at 10.02.32

Screen Shot 2020-03-14 at 10.02.45
Source: Wikipedia
110620383_spread_of_sars_640_v2-nc
Source: The Globalist

The pattern does look different for the seasonal flu.  Unlike temperate climates, where flu actually is seasonal, flu cases are reported year-round in many African countries.  Surveillance data on the flu in African countries is relatively weak and there are few country-specific flu vaccines available.  It’s possible that cases of the flu are often misdiagnosed as malaria, which is also highly prevalent.

Seasonal flu (2011 – 2016)

Global-maps-of-monthly-influenza-activity-2011-2016
Source: Newman et al. (2018)

Let’s look at the pre-1980 epidemics now.  These include outbreaks of H1N1 / Russian flu (1977), H3N2 / Hong Kong flu (1968), H2N2 / Asian flu (1957), and H1N1 / Spanish flu (1918).  I couldn’t find a good map for the Russian flu.  For the other epidemics, I could only find maps tracing the path of the outbreak, rather than displaying the number of cases.

The 1968 Hong Kong flu looks fairly similar to the transmission path for coronavirus, in that Latin America and Africa seem to have been minimally affected.  However, the 1957 Asian flu and the 1918 Spanish flu did impact Africa.  I’m not sure about the mechanism by which the 1957 flu spread, but in 1918 the flu was spread by some of the two million African soldiers who returned home after being forced to fight for their colonizers during WWI.  African countries had some of the highest mortality rates in the pandemic, with 2% of all Africans dying of the flu in only six months.

hong kong flu
Source: Rybicki (2015), via Twitter
Asian_Flu_Map-large
Source: Johns Hopkins School of Public Health

H1N1 / Spanish flu (1918)

spanish flu
Source: Thinglink

There are many reasons why we can’t conclude very much about the spread of the novel coronavirus in Africa based on patterns of previous outbreaks.  Transportation patterns, public health capacity, and the nature of the disease itself are all different for each pandemic.  It’s still important for African countries to continue preparing their health systems for a more widespread coronavirus outbreak.

The DRC has hit an Ebola breakthrough

 

ebola 2
Ebola responders at work, via Al Jazeera

For the first time since June 2018, the DRC has reported that it no longer has any known cases of Ebola.  This comes 14 days after the last new case of Ebola was reported.  If the country makes it another 28 days without a new case, it can officially declare the epidemic over.

Of course, one key challenge is that it’s hard to do surveillance in remote rural areas, many of which are under the control of various militias.  So while this is a reason for optimism, it’s far too soon to declare that the epidemic will be over soon.