Global government debt surged to nearly 100 percent of GDP during the global recession of 2020, as the COVID-19 pandemic triggered a collapse in output and governments provided unprecedented fiscal support. As the global economy recovers and fiscal support is withdrawn, a key question is whether government debt (relative to GDP) will stabilize and start to decrease. In a new study, we answer this question by analyzing the evolution of government debt after previous recessions.

Government debt: often increases after global recessions

Historically, global government debt has increased after every global recession over the past six decades. Between 1960 and 2019, there were four global recessions: 1975, 1982, 1991, and 2009. Global government debt rose by a cumulative 4-15 percentage points of GDP over the five years following these global recessions—by 4 percentage points of GDP over 1975-80, 15 percentage points over 1982-87, 9 percentage points over 1991-96, and 4 percentage points over 2009-14 (Figure 1).

Government debt also tended to be higher after recessions in a majority of countries. On average, in the five years after a global recession, two-thirds of countries had the same or higher debt levels. A slightly larger share of advanced economies saw higher levels of debt after recessions than emerging market and developing economies (EMDEs), while around three-quarters of low-income countries (LICs) had higher debt after recessions.

Advanced economy debt has seen a consistent jump in the five years after every global recession, with an increase of 3-14 percentage points after the global recessions prior to 2020 (Figure 2). The last three recessions all saw an increase in advanced economy debt of more than 10 percent of GDP.

In contrast, the evolution of government debt in EMDEs has been more erratic (Figure 3). Government debt in EMDEs excluding China saw small declines in the five years after the 1991 and 2009 recessions. For the 1991 recession, debt rose in the immediate aftermath of the recession but then decreased rapidly as growth recovered. While in the 2009 recession government debt saw a modest increase during the recession but stabilized thereafter, as EMDEs were less affected and recovered more rapidly from the global financial crisis than advanced economies.

Regional dimensions: a mixed picture but not for everyone

Regionally, the evolution of government debt after global recessions was more varied. Almost all regions saw an increase in debt following the first two recessions, with particularly large increases in East Asia and the Pacific (EAP), Latin America and the Caribbean (LAC), and sub-Saharan Africa (SSA). EAP and LAC saw an unwinding of this debt in the period after the 1991 recession as debt was reduced, including due to the provision of debt relief (via the issuance of Brady bonds), while debt in SSA rose further as many countries did not receive debt relief until the late 1990s.

Debt was broadly stable in most regions following the 2009 recession, which mainly affected advanced economies. Overall, all regions other than SSA saw at least one global recession episode in which government debt declined. In SSA, however, government debt has increased following each previous recession, and debt only declined during the late 1990s and 2000s as a result of the Heavily Indebted Poor Countries Initiative and Multilateral Debt Relief Initiative. SSA has the largest number of LICs, and more than half of LICs are in debt distress or at high risk of debt distress.

Hopes and realities: no time for complacency

In the medium term, some expect that global government debt stocks will stabilize at current levels as a result of the post-pandemic rebound in growth and withdrawal of fiscal support measures. The expected stabilization in debt-to-GDP ratios may alleviate some concerns about elevated debt levels at present.

If such a stabilization materialized, however, it would be a significant departure from debt developments in the aftermath of previous recessions, particularly in the case of advanced economies and countries in SSA. Furthermore, forecasts of government debt tend to suffer from optimism bias: actual government debt to GDP ratios have been shown to be about 10 percentage points of GDP higher after five years than initially forecast, on average.

In light of this historical record, and given large financing gaps and significant investment needs, including facilitating the energy transition, a stabilization in debt levels looks optimistic. Even if debt does stabilize, it remains at exceptionally elevated levels by historical standards and may rise if current low interest rates do not persist.

What are the implications of these observations for policymakers? It is critical to avoid complacency among policymakers who may have optimistic views about debt prospects in the near term. Some policymakers in EMDEs may be tempted to rely on growth alone to lower debt while some others hope that low interest rates would help keep debt service manageable. Policymakers should hope for the best but prepare for the worst as a new monetary policy tightening cycle gets underway in advanced economies.

      

Starting in 2010, a new wave of debt accumulation—the “fourth wave” of debt—had been underway in emerging market and developing economies (EMDEs, Figure 1). With the sharp increase in debt during the COVID-19 pandemic, the fourth wave of debt has turned into a tsunami and become even more dangerous. The tsunami of debt has amplified the difficulty of resolving debt not just because of record debt levels but also because of significant changes in the structure of debt markets.

Sharp and broad-based increase in debt

In 2020, total global debt rose by 30 percentage points of GDP, to 263 percent of GDP—the largest single-year increase since at least 1970 (Figure 2). This increase was broad-based, evident across government and private debt, domestic and external debt, and the majority of countries. In EMDEs, total debt went over 200 percent of GDP, and in advanced economies, total debt exceeded 300 percent of GDP in 2020.

As output plummeted in the worst global recession since World War II and governments enacted unprecedented fiscal support measures, global government debt registered its fastest single-year jump to roughly 100 percent of GDP, its highest level in half a century. The increase was near-universal. Government debt rose in almost nine-tenths of countries and at its fastest pace in half a century in around one-quarter of countries.

Private debt also rose at a record pace and to an unprecedented high in 2020 as output collapsed, lockdowns closed businesses, and fiscal, monetary, and regulatory policy measures supported credit extension. Globally, it jumped by 15 percentage points of GDP to 165 percent of GDP in 2020, its highest level since records started in 1970. Similar to government debt, the jump in private debt was broad-based, affecting more than four-fifths of countries.

The bulk of the increase in total debt in 2020 was accounted for by rising domestic debt. In advanced economies, rising domestic debt accounted for about one-half of the increase in total debt in 2020; in EMDEs, for nine-tenths; and globally for three-fifths. In advanced economies, less than one-half of total debt (and less than one-third of private debt) is domestic whereas, in EMDEs, total debt remains predominantly domestic (85 percent of total debt), especially for private debt, which is nine-tenths domestic. Global external debt also rose by 12 percentage points of GDP in 2020, to 114 percent of GDP in 2020, and, as a result, the stock of external debt at the global level now exceeds its 2010 level, after a decade of decline.

Larger risks

Thanks to exceptional monetary policy support, borrowing costs have fallen to record low levels. Low global interest rates can make additional debt-financed government support seem attractive, especially when growth rates are expected to be above interest rates. However, rapid debt accumulation, triggered by fiscal support measures, still entails significant risks. There is no guarantee that any future economic shock will lower interest rates.

In addition, EMDEs face a risk of sudden stops of capital flows, especially if they have other vulnerabilities like sizeable foreign currency debt, overvalued exchange rates, and financial system fragilities. This is also recognized in sovereign risk ratings: The fraction of EMDEs that experienced sovereign credit downgrades in 2020 was larger than that over the 2010-19 period, whereas in advanced economies it was smaller (Figure 3).

Risks associated with rapid debt accumulation have been magnified because of the likely adverse impact of some policy interventions on policy frameworks. Many governments have heavily encouraged credit extension and eased regulatory policies since the beginning of the pandemic. These necessary measures prevented a credit crunch. However, private sector liabilities could eventually migrate onto government balance sheets, either in a financial crisis or, indirectly, in a protracted period of low growth.

Similarly, unprecedented monetary policy support measures were necessary due to the magnitude of the pandemic shocks. However, these policies may also sow the seeds of solvency problems that may become apparent once global interest rates begin to rise from historically low levels. The sharp increase in debt in a short period and looser fiscal controls also heighten the risk that not all of the debt was used for productive purposes.

New policy challenges

For countries that get into debt distress, achieving a successful resolution may be more difficult than it was in the past. Specifically, future debt restructurings will likely be more complicated because of a more fragmented creditor base than in the past and a lack of transparency in debt reporting. The importance of bilateral non-Paris Club lenders has increased significantly over the past decade (Figure 4). The increasing number of private creditors and range of financial instruments further complicates debt resolution. The growing diversity of creditors and complexity of debt instruments has been associated with greater uncertainty about the level and composition of debt, as not all creditors are bound by a single set of reporting standards and loan terms are often confidential.

In several dimensions, the playing field is currently tilted in favor of creditors and discourages prompt and comprehensive debt resolution. While 91 percent of sovereign bond issuance since 2014 has included collective action clauses that facilitate restructuring, a large legacy stock without such clauses remains: about 50 percent of outstanding international debt does not include collective action clauses. In the past, delays in resolving high debt were associated with weaker output and investment growth, raising the prospect that even slower debt resolution could lead to a lost decade of growth in some countries already facing significant debt problems. In addition to sovereign debt, the rise of private debt, especially during the pandemic, highlights the importance of effective resolution frameworks to mitigate the risk of private debt overhang.

Urgent action needed

The challenges of resolving record-high debt point to the urgency to act on the parts of both national policymakers and the global community. National policymakers will need to improve policy frameworks to make debt sustainable, as well as to consider the best approaches to resolving debt if it becomes unsustainable. The global community needs to act rapidly and forcefully to ensure that this debt tsunami does not end with a string of debt crises in EMDEs, as earlier waves did.

      

If the international community is to meet global development and climate challenges, in particular the Sustainable Development Goals (SDGs) and the Paris Agreement climate targets, everyone involved must ask themselves how their innovations, projects, or activities can identify successful interventions and scale them up. Over the last 15 years, teams at Brookings have tried to figure this out by taking stock of what we know from literature and practice, by exploring how to bring development solutions to millions of poor people, by looking at scaling opportunities in agriculture, education, and climate change and, most recently, by studying how breakthrough innovations can help achieve critical development outcomes at scale.

We also looked at whether and how development finance agencies—multilateral, bilateral, foundations, and international NGOs—scale up systematically by hardwiring these good habits into their funding and operational approaches. We found that traditional development finance focused on time-bound, one-off projects to deliver limited results. When these results were achieved, the projects were rated “successful” in ex-post evaluations. Scalability and scaling after a project ends have not featured as a criterion for successful project delivery in project design, implementation, or evaluation. As a result, too many development projects have been “pilots to nowhere,” so their benefits end up being limited in size and duration.

Donor organizations can help recipients to scale in many ways, not only as funders, but also as champions, capacity builders, intermediaries, and coordinators. And demonstrating impact at scale can help donors convince the public, parliaments, and politicians that they deserve to be funded from scarce taxpayer resources. At the same time, one-off donor projects that do not scale tend to reinforce the focus on short-term impact among local partners, whether in the public or the private sectors. So, an effective, systematic approach to scaling by donor organizations is critical.

5 stages

How should one go about changing prevailing donor practice? A first step is to realistically appraise the status of scaling in a donor organization. To do so, a range of options can be considered.

Going from left to right, we can distinguish five types of integration of scaling into an organization:

• The scaling agenda is pursued outside the organization by funders, consulting outfits, think tanks, researchers, or advisers trying to support and influence the organization in mainstreaming scaling.
• Scaling is the focus of a separate unit at the periphery of the organization, but not integrated into operations of the “mother ship”; this is typical for innovation and accelerator labs introduced by many development finance organizations in recent years (e.g., DFID, UNDP, WFP).
• Scaling is pursued somewhere in the organization by one or more individuals or units, but not all; this appears to be the case now in such organizations as the World Bank and USAID.
• The scaling agenda is at the center of the organization, i.e., it is owned by the organization’s leadership and is part of its mission and strategy but is not fully integrated into operational practices. The CGIAR (the Consultative Group of International Agricultural Research) and the GCF (Green Climate Fund) recently put scaling at the center of their strategies; they are still in the process of mainstreaming it throughout the organization.
• Finally, scaling is intrinsic to the organization or fully mainstreamed, with scaling systematically pursued in operational practice throughout the organization; examples where this has been achieved include some of the specialized funding institutions (“vertical funds”), such as The Global Fund and the newly established Systematic Observations Financing Facility (SOFF), and some of the smaller foundations, e.g., the Eleanor Crook Foundation.

Typically, there’s been a concentration of organizations in the left three circles, with a few outliers to the right. Over recent years, there has been some progress from left to right, but there are still few cases of full mainstreaming.

6 steps to scale

In attempting to move an organization from left to right in the graph above one needs to develop a systematic approach to mainstreaming scaling consisting of six steps:

1. Create a shared vision, e.g., as part of a mission statement; this is an important first step but by itself not nearly enough.
2. Develop a systematic approach to scaling that is appropriate for the organization and its mission, by assessing and adapting current organizational practices in five areas: (a) vision and strategy; (b) operational business models and instruments; (c) internal policies, guidelines, and management processes; (d) staff and management; and (e) monitoring and evaluation. In the process one must adapt the standard one-off project model to incorporate a systematic focus on sustainable scaling beyond project end.
3. Strengthen enablers for mainstreaming (leadership from the top, champions, managerial and staff incentives, partnerships, etc.).
4. Reduce barriers (bureaucratic inertia, staff overload with “new fads and buzz words,” budget constraints on staff training, etc.).
5. Take appropriate action and follow through.
6. Monitor and evaluate progress with mainstreaming.

Like all change management and scaling efforts, mainstreaming scaling into an organization takes time and persistence and faces risks of discontinuity (especially with changes in leadership).

The International Fund for Agriculture and Development (IFAD) followed this approach with the assistance of a team of experts from the Brookings Institution and other external advisers over the last 10 years. An independent evaluation showed that much had been achieved, but that additional efforts were required for full mainstreaming. IFAD is now updating its scaling strategy.

In pressing forward with mainstreaming scaling in a donor organization one has to expect pushback from those interested in maintaining the status quo. Typical objections include these:

• We’re big and already doing it (for large organizations).
• We’re too small (for small organizations).
• It’s yet another fad best ignored.
• My management/board/authorizing environment only care about project impact, not about longer-term impact at scale.
• We don’t have the budget; our front-line is already overloaded.
• Your methods/operational guidance are too generic (or complex and burdensome).
• We can’t (be bothered to) keep track of whether the scaling approaches are implemented and working as intended.

10 lessons

From my work with some 15 international development agencies and funds, I’ve learned the following lessons:

1. Focus on changing the mindsets of managers and staff by showing them that a systematic approach to scaling will help them achieve shared development objectives (e.g., SDGs) more effectively.
2. Remember that scaling is about development impact, not necessarily about more money or bigger organizational footprint.
3. Develop an approach that responds to the specific needs of the organization (with own managers/staff in the lead, not outsiders).
4. Focus on integrating scaling into the project cycle, with a special focus on what happens beyond project end.
5. Provide practical guidance and training to staff (and consultants).
6. Keep bureaucratic processes as simple as possible.
7. Allow more risk taking, recognizing that lack of success in projects is an opportunity to learn, rather than a failure.
8. Link up with clients, partners, and funders in pursuing the mainstreaming agenda.
9. Monitor and evaluate the mainstreaming process.
10. Stick with it! The mainstreaming process takes years to complete.

      

It is shopping time. Many people, especially in the Northern Hemisphere, participated in Black Friday and Cyber Monday while getting ready for Christmas. For many, the 2021 shopping experience also reflects a return to normality at a time of never-ending waves of COVID-19 infections and lockdowns.

Many large stores are in cities, which reflect two megatrends that are reshaping the global economy: the rise of the global consumer class and urbanization. Both megatrends experienced tipping points in this century. In 2008, the world became majority urban and in 2019, World Data Lab projected that half the world would be middle class or wealthier. By the end of 2021, there will be 4 billion people in the global consumer class, and, in the absence of another major economic crisis, the global consumer class will reach 5.2 billion people by 2030.

Urbanization will also continue steadily throughout this decade. People are moving from villages to cities to find better education, health care, and jobs. Additionally, in emerging markets, former villages are contributing to urbanization as they grow rapidly to the point that they represent a peri-urban segment. These are often the new suburbs of the world’s megacities such as Jakarta, Mexico City, Mumbai, São Paolo, or Lagos.

Urban areas are more prosperous than rural areas. Proximity breeds innovation and allows for economies of scale. In “Triumph of the City,” Edward Glaeser showed that cities are healthier, wealthier, and better for the environment because higher population density makes it possible to produce new goods and services at scale and thus at a lower cost for everyone. This is also why most of the global consumer class is now living in cities.

To estimate the urban consumer class, we use urbanization projections from the U.N. along with education and income measures and investigate how personal spending depends on each of these factors, especially urbanization. Projections for India are derived from survey microdata for the country.

World Data Lab projects that out of today’s 3.8 billion consumers, almost 3.2 billion (80 percent) are living in urban areas (see Table 1).

Table 1. Two-thirds of the global consumer class are urban

Source:  MarketPro, World Data Lab projections.
Note: Consumer class = greater than $11/per day. Poor and vulnerable = less than $11/per day.

However, the split is uneven across countries as many emerging markets are still predominantly rural. In OECD countries, almost everyone is part of the consumer class. Since OECD economies are highly urbanized, approximately 80 percent of the OECD consumer class is urban. The same is true for emerging markets and developing countries but for different reasons. Poorer countries are still majority rural, but only a very small portion of rural folk are part of the consumer class. This means that among poorer economies also approximately 80 percent of consumers are in urban areas. In Asia where urbanization has been advancing rapidly, the urban consumer class dominates with 1.7 billion people, representing approximately 54 percent of the global urban consumer class. Asia’s consumer class is not only rising in China and India. There are also several emerging markets that are below the radar, especially Indonesia, Bangladesh, Pakistan, the Philippines, and Vietnam.

But India remains the most dynamic market for urban consumers because it remains relatively rural even though it is urbanizing rapidly.  In India both megatrends converge on scale. Today India has an estimated 208 million urban consumers, which will rise to an estimated 374 million. This represents an annual growth of 6 percent. The growth of India’s urban consumer class is six times larger than global population growth, almost double growth of the global consumer class, and even substantially larger than the average of urban consumer class growth and Asia’s consumer class growth (see Figure 1).

Figure 1. The consumer class growth pyramid—India is on top

Source: World Data Lab projections.

Many of us will enjoy the festive season of the year looking back at two difficult years dominated by the COVID-19 pandemic. Despite the massive economic impact of COVID-19 in 2020, its impact has only been temporary. The fundamental forces of global megatrends will shape this decade and they will be most visible in Asia.

For questions on the underlying data model, please contact Max Heinze (max.heinze@worlddata.io).

      

China’s rapid economic progress has been associated with a rapidly rising carbon footprint. The country now wonders how it can extend its economic bull run while achieving its ambitious climate goals. In this blog, we look at China’s climate challenge through the lens of urban planning. By embracing more populous and denser megacities, China could achieve higher economic efficiency and lower emissions at the same time.

Denser Chinese cities have lower per capita emissions

In many developing countries, urbanization is often associated with higher carbon emissions, resulting from rising consumption levels and associated energy needs. However, beyond a certain level of per capita income, as people move into denser urban spaces, their carbon footprint declines. The recent World Bank Commodity Markets Outlook shows that urban density tends to lower energy consumption intensity from transport, infrastructure, and accommodation. Researchers at Australian National University have found that this is true also for China. As illustrated in Figure 1, there is a statistically significant negative relationship between population density and per capita emissions, even after controlling for income, economic structure, and proxies for environmental policy.

Figure 1. Population density of Chinese cities is negatively correlated with per capita emissions
Log of CO2 emissions per capita and log of population density for 175 Chinese cities

Source: World Bank calculations using CEAD emissions data for 175 prefecture-level and above cities in 2010 and population density data from Liu et al. (2020).
Note: Population density is measured in 10,000 people per square kilometer. CO2 emissions per capita are measured in 100 tons per person.

This finding is important because China is still urbanizing. Assuming an urbanization rate of 70 percent by 2030, some 80 million people may move to urban areas over the next decade alone. For China’s economy and its emission intensity, it matters where these people move to.

In the past few years, many of China’s megacities, such as Beijing and Shanghai, have made plans to cap the size of their populations. China has traditionally controlled rural-urban migration through the hukou household registration system. In recent years, such controls have been lifted in smaller cities but tightened in megacities. Meanwhile, Chinese cities have been expanding outward: China is the only country in East Asia to have declining population density in its large cities, as the expansion of urban boundaries has outpaced population inflows. This does not augur well for efficiency or emissions.

An anomalous city-size distribution

A striking pattern that has long fascinated both urban scientists and economists is that the distribution of city population size and rank in many countries seems to closely follow the rank-size rule. That is, the rank r associated with a city of size S is proportional to S to some negative power. The special case in which the estimated power equals –1 is known as Zipf’s law, named after a linguist, George Zipf. Of course, this empirical regularity serves mainly illustrative purposes. Benchmarking China’s city size distribution against Zipf’s law reveals how government policy has shaped China’s urbanization in ways that may run counter to economic forces.

In the early 19th century, China’s city size distribution closely followed the rank-size rule. After the 1960s it diverged substantially, as industries and industrial centers were relocated from the coast to the interior under the Third Front Movement. Since the mid-1990s, as migration to the coastal areas increased, China’s largest coastal cities started to grow rapidly, restoring a more natural pattern of urbanization. Nonetheless, China’s megacities remain significantly less populous and less dense than one might expect, and its mid-size cities are much larger (Figure 2). Indeed, if Zipf’s law held, Beijing and Shanghai would be respectively three and four times more populous. Zipf’s law is a special case that tends not to hold in reality, but even a more moderate rank-size rule would imply substantially more populous megacities. If Guangzhou had the same density as Seoul, for example, it could accommodate 4.2 million more people.

Figure 2. If Zipf’s law held, China’s megacities would have far higher populations
Current population size and rank, current vs. implied by Zipf’s law

a) In logs

b) In levels

c) Current population density 

d) Density if Zipf’s law held

Source: World Bank calculations using CEAD emissions data for 175 cities in 2010 and population density data from Liu et al. (2020).
Note: Population is measured in 10,000 people. This implied Zipf’s law distribution is taken holding the 95th percentile city’s size and rank fixed.

One reason for this is that China’s urban planners have discouraged its megacities from growing too large in population size. Resistance by existing residents who fear the influx of rural migrants and the overcrowding of public services may be one reason. China’s fiscal system, in which local governments are highly dependent on land sales to raise revenues, has also contributed to an extensive model of urbanization, lowering population density.

Denser megacities would be good for emissions, and high productivity services

Allowing China’s prosperous coastal cities to grow further in size and to become denser in the process could play an important role in decarbonization.  A back-of-the-envelope calculation suggests that population inflows into China’s largest cities in line with Zipf’s law, for example, could be associated with declining per capita emissions in the order of 7 percent on average for China’s 10 largest cities.

These emission declines would come at little cost—indeed, more densely populated cities would likely also be more productive. The recent World Bank Pancakes to Pyramids report shows how smart urban planning, combined with investments in infrastructure, can reap the economic benefits of denser cities while turning them into vibrant livable spaces that are attractive to the best talents. Concentrating more jobs and economic activities in China’s most productive cities could do much to help shift China’s growth model toward innovation, domestic consumption, and high value services.

In 2014, the World Bank’s Urban China report stressed the importance of city density and labor mobility in China’s next development phase. China’s new climate goals add another reason why this agenda remains more relevant than ever.

      

The ongoing policy discussions about climate change resilience and the role of prevention and insurance, including the recent blog by Matthew Kahn and Somik Lall, reminded me of the 1985 science fiction movie “Back to the Future,” where a teenager is accidentally sent three decades back in a time-traveling automobile built by his eccentric scientist friend “Doc” Brown. It may be indeed worthwhile to go back to the future and return to the foundations of the economics of risk and insurance because some fundamental concepts developed more than half a century ago may be still very relevant to inform policy actions on climate change resilience.

The 50-year-old economic framework can offer some forward-looking insights on how individuals or households behave when confronted with climate shocks, the interactions between risk prevention and market insurance, and the market-enhancing role of government, especially when insurance markets are incomplete.

One of those seminal economic papers was published in 1972 by Isaac Ehrlich and Nobel laureate Gary Becker. In it they analyze how a risk-averse individual with a “risky” asset or wealth, such as a house exposed to floods, should determine their optimal mix of preventive investments that allows them to (i) reduce the magnitude of their potential loss (self-insurance), and/or (ii) reduce the probability of their loss occurring in the first place (self-protection) and/or (iii) purchase market insurance. In the context of climate change, self-insurance investments include building houses on stilts to protect them against floods, self-protection investments include migrating away from disaster-prone areas, and market insurance includes catastrophe risk insurance programs.

The individual decisions on optimal preventive investments to protect a risky asset can be reexamined using the concept of “riskiness” defined by Michael Rothschild and Nobel laureate Joseph Stiglitz in their 1970 seminal paper on increasing risk. Using the Rothschild-Stiglitz definition of “riskiness,” self-insurance and market insurance unambiguously reduce the “riskiness” of the asset by transferring wealth from good states of the world (when no loss occurs) to bad states of the world (when a loss occurs). On the contrary, an increase in self-protection investment may increase the “riskiness” of the asset: Self-protection can make a risk-averse individual worse off if the disaster loss still occurs because the individual not only experiences the disaster loss but has also paid a higher cost for self-protection. The concept of “riskiness” also explains why more risk-averse individuals invest more in self-insurance and market insurance but may invest less in self-protection.

The interactions between those preventive investments can be reexamined in light of their impact on the risky asset and allow us to re-interpret the two fundamental results of Ehrlich and Becker: Market insurance and self-insurance are substitute because they both contribute to reducing the “riskiness” of the asset; their optimal combination depends on their respective (shadow) price. This means that an increase in insurance coverage, for example, as a result of insurance premium subsidies, will reduce investments in self-insurance. On the contrary, market-insurance and self-protection can be complements: While market insurance makes the asset less risky, self-protection can increase its “riskiness.” This means that an increase in insurance coverage may induce increased investments in self-protection.

But what if the insurance market is incomplete? Let’s assume for example that there is a (uninsurable) risk that the insurance company does not pay its claims. This risk of insurance being nonreliable may be because the insurance company goes bankrupt or because of litigation with the settlement of the claims. During the COVID-19 pandemics for example, businesses and insurers went to court battles over whether insurers have to pay out claims over business interruptions as a result of the pandemic. Using Rothschild-Stiglitz’s definition of “riskiness,” nonreliable insurance does not always reduce the “riskiness” of the asset: The insured individual is worse off when the disaster loss occurs because she does not receive her insurance payouts while she had paid her insurance premium. This means that, when the insurance market is incomplete, self-insurance and nonreliable insurance can be complements. This potential complementarity means that an increase in (nonreliable) insurance coverage may lead to an increase in self-insurance investments. In other words, the insured individual may over-invest in self-insurance to mitigate the risk of market insurance being nonreliable. This is an inefficient use of self-insurance caused by insurance market imperfections.

Figure 1. Self-insurance and (non)reliable insurance as substitutes (complements)

Source: Author.

This simple framework that builds on economic concepts developed more than five decades ago can provide some insights for public policy to strengthen financial markets for improved climate change resilience when insurance markets are incomplete. The complementarity between self-insurance and nonreliable insurance is clearly an inefficient allocation of resources due to incomplete insurance markets. There is therefore a rationale for public intervention and policy reforms to help correct these market imperfections. Prudential regulation of insurance markets, for example through solvency capital requirements that account for climate risks and other systemic risks like pandemics, can help improve the soundness of insurance companies’ operations. Likewise, the establishment of (government-sponsored) insurance guarantee schemes can protect the insureds when insurers are unable to fulfill their contractual commitments. Resilient financial markets will ensure that both market insurance and self-insurance, as substitute investments, maximize their synergy and contribute to improving the resilience of countries, communities, and people against climate shocks.

      

Land tenure—the formal and informal relationship individuals and groups form with land—effectively determines who uses what land under which conditions. Tenure security is important to promote rural resilience and climate change adaptation, build endowments of assets, and provide adequate housing. But land tenure security is not static. It exists along a spectrum from strong to weak with numerous factors moving the needle between the two.

One such factor is conflict and all the elements that comprise it. Violence, forced displacement, land grabbing, destruction of land records, death of land, and weakened governance associated with conflict affect tenure security and people’s access to land. Understanding exactly how rights are affected can help design interventions to enhance land tenure security during and after periods of conflict.

One way in which land tenure is impacted by conflict is through people’s perceptions of the security of their tenure under the cloud of conflict and the risks it brings. In 2020, Prindex, through its global survey, released data on perceptions of security of land and housing rights in 140 countries. Comparing this with data from the Uppsala Conflict Data Program’s georeferenced records of violent events sheds light on the relationship between conflict and tenure security.

The key findings:

1. There is a clear link between conflict, fragile settings, and perceived tenure insecurity. Countries with medium- to high-intensity conflict or institutional/social fragility tend to have relatively high rates of perceived tenure insecurity. Countries with a record of violent events in the past also have higher rates of perceived tenure insecurity. Experiencing at least one event of organized violence in the previous year pushes a country’s overall rate of tenure security down by about 10 percentage points on average (71 percent vs. 81 percent).

2. Renters are more vulnerable to negative impacts of violent conflict. Conflict negatively impacts security of all three types of tenure measured—ownership, renting, and living with family. The most negatively affected are renters, 67 percent of whom feel secure in stable countries versus only 50 percent in countries that have experienced organized violence.

3. Violence acts as a “leveler” in terms of the effect of income on perceived tenure security. Rich or poor, organized violence makes people feel more insecure about their land rights. Overall, low-income households are the most insecure in violent and nonviolent settings. However, violence reduces perceived tenure security much more sharply among high earners than low earners. One possible explanation is that lower-income households may face tenure insecurity on multiple fronts, while for higher-income households organized violence presents a unique shock.

4. Organized violence pushes people into less secure housing. Violence seems to push people from more secure arrangements, such as living with family, into less secure ones, such as renting. The presence of organized violence is associated with an increase in the share of renters by 2.7 percentage points and a decrease in the share of those staying with family by 5.7 percentage points over a 28-year period. Violence in rural areas can spur increased urbanization, presenting risks to vulnerable persons renting urban housing, as evidenced in a new World Bank report on Somalia.
5. Organized violence has a long-term effect on tenure security. Organized violence affects perceived tenure security among the general population for as long as 28 years. The magnitude of the effect is comparable to the effects of lower levels of education or low income on perceived tenure security. This effect is considerable and is likely to have negative development impacts, especially where it prevents people from using land as an asset.
6. The relationship between the severity of organized violence and perceived tenure security is less clear. The number of deaths or events revealed no more insights than binary measures of violence versus nonviolence.

A better understanding of the relationships between violence and perceptions of tenure security will help develop targeted policies and interventions that safeguard and strengthen tenure security during and after conflict. Additional research on how additional factors, such as the length of conflict, causes (e.g., expropriation vs. family disputes) and spillover effects into non-conflict areas, impact tenure security would help further fill knowledge gaps. Recognizing and addressing the triggers of perceived insecurity could also help governments build trust in post-conflict settings, a critical element in promoting peace and stability, given the social and economic significance of land and housing in conflict and fragile settings.

      

The labor market is the main vehicle through which the proceeds of growth are shared among households and individuals. Therefore, understanding the labor market is essential for poverty reduction. This topic is crucial in Nigeria, where the government aspires to lift 100 million Nigerians out of poverty by 2030—an ambitious objective since, even before the pandemic, around 4 in 10 Nigerians lived below the national poverty line.

COVID-19’s “double shock”—health and economic—has intensified the need for new evidence to understand jobs and livelihoods in Nigeria. With social protection limited, households resorted to negative coping strategies—including reducing food consumption—that hurt their current and future welfare.

A new report, COVID-19 in Nigeria: Frontline Data and Pathways for Policy, uses high-frequency data to examine effects on human capital, livelihoods, and welfare. The report draws on the Nigeria COVID-19 National Longitudinal Phone Survey (NLPS), a distinctive, nationally-representative survey that captured key socioeconomic information from households for 12 consecutive rounds between April 2020 and April 2021.

Employment during COVID-19: Quick drop, quick recovery

The NLPS data show that employment in Nigeria plummeted at the start of the COVID-19 crisis. The share of main respondents in each household who were working fell by more than half between mid-March 2020 and April/May 2020, dropping from 86 percent to 42 percent (Figure 1). During this period, the most stringent lockdown measures were in place, and restrictions on mobility may have stopped people from getting to work. This also chimes strongly with global evidence from other labor markets.

Despite the initial drop, employment in Nigeria recovered quickly. By August 2020, the share of main respondents in each household who were working had returned to pre-pandemic levels. In this sense, Nigeria’s labor market echoed the V-shaped recovery observers hoped for in the global economy as a whole, following the COVID-19 crisis.

A closer look: Many people working, but not in good jobs

However, looking at the types of jobs Nigerians turned to paints a less positive picture. Later rounds of the NLPS—those implemented in September 2020 and February 2021—expanded interviews to all working-age household members, capturing more detailed and inclusive information on Nigeria’s labor market. The share of working-age Nigerians who were working actually increased between January-February 2019 and February 2021, but this was mainly concentrated in retail and trade (or commerce) activities in non-farm household enterprises. Such activities are typically small-scale—with only around 1 in 10 non-farm household enterprises employing anyone outside the household—so would be unlikely to help households ward off or escape poverty. Indeed, NLPS data directly demonstrate that non-farm enterprise income remained the most precarious—more so than wage work or agriculture—as  COVID-19 continued.

Moreover, COVID-19 heralded significant churn in Nigerians’ labor market activities. Workers lacked stability and security in their employment: Instead, they took on whatever activities could help them cope with the effects of the COVID-19.

Learning losses place future growth at risk

Given its impacts on human capital development, and especially education, the crisis also threatens future generations. School closures during 2020 reduced children’s attendance rates even after reopening, especially among older children. Dropout was also higher in the households most affected by income shocks, suggesting that households removed children from school in order to support income-generating activities. Since Nigeria’s human capital outcomes were well below the average for sub-Saharan Africa even before the pandemic, the country can ill afford these setbacks to learning.

COVID-19 also threatens to widen inequality in learning, as access to remote learning was uneven across households. Young children from non-poor households had better access to remote learning options—through television, computers, and smartphones or tablets—than those from poor households (Figure 2).

A window for policy action

Recouping the learning lost during the COVID-19 crisis, therefore, presents a key policy priority for Nigeria. While encouraging children back to school—the preferred policy among Nigerians themselves—will be vital, resuming in-person learning requires that preventative measures be in place to prevent the virus’ spread. With ongoing uncertainty about the path of the pandemic, remote options that actually work for the poor are needed. High-tech options cannot reach the poor, so low-tech solutions may be more appropriate. Examples include engaging pupils, parents, and teachers through mobile phones or broadcasting lessons via radio. Further initiatives could support the recovery in learning, be it in person or remote: For example, there is growing evidence that Teaching at the Right Level (TaRL) can support foundational learning by carefully assessing children’s needs and then tailoring teaching accordingly.

The crisis also provides renewed impetus to implement the policies needed for good job creation in Nigeria. As well as investing in human capital, this partly hinges on effecting macroeconomic reforms to energize structural transformation and generate productive wage jobs. Yet, since farm and non-farm household enterprises will dominate employment in Nigeria for many years to come, policies to boost their productivity—through developing crop varieties, investing in infrastructure, improving market access, and easing credit constraints—should also be carefully considered.

The country’s large youth population makes it even more vital that Nigeria’s leaders apply evidence-based policies to exit the crisis and support the country’s workers, today and tomorrow.

      

Negotiations at COP26 focused on green technology and finance. Governments pledged money, businesses committed to net-zero production, and ordinary citizens … did nothing! Individual activists made a lot of noise, but there was no systematic effort to organize the change in consumption patterns needed to reach our shared goal of keeping climate warming to less than 1.5 degrees Celsius. The Chichester Festival Theater organized a crowd of eco-activists to spell out “commit,” and to pledge to reduce food waste and the like, but the lifestyle and behavior changes of individuals, especially middle-class and rich consumers, received far less attention than warranted. The richest 10 percent of consumers account for 44 percent of consumption-related carbon emissions.

A few facts can provide context. About two-thirds of global greenhouse gas (GHG) emissions are linked to household consumption. This is why the U.N. Environment Program’s (UNEP) 2020 Emissions Gap Report concluded that major lifestyle changes will be required. Consumers will need to reduce their carbon footprint from a global average of around 6 tons of CO2 equivalent (C02eq) per person to 2-2.5 tons by 2030 and to 0.7 tons by 2050. Some of this is done automatically when businesses produce in more sustainable ways. For example, when utility companies substitute renewable sources for fossil fuels, the indirect emissions of consumers in heating and cooling their homes using electricity automatically decline. The consumer is not being asked to do anything except, perhaps, to switch to electric appliances. They can retain their consumption pattern. But this will not suffice. Changes in lifestyles are needed. That’s why Sustainable Development Goal 12 is responsible consumption and production.

The richest 10 percent of consumers account for 44 percent of consumption-related carbon emissions.

Every little bit counts given the scale and urgency of reducing emissions, but where consumers are concerned there is a flood of suggestions and recommendations that generate more confusion than actionable information. Research on how consumer lifestyle choices are affecting aggregate carbon emissions has lagged.

We know something about the differences between countries. The average consumer in the United States, for example, emits about 17.6 tons of CO2eq per capita, more than double that of the European Union and the U.K. (7.9 tons), and 10 times as much as India (1.7 tons). What we don’t know with any degree of robustness is how much this is simply due to higher income and spending levels in the U.S. (rich people emit more than poor people), how much is due to temperature and other natural conditions, and how much is due to policy choices.

A just transition would take into account all these issues (and unsurprisingly, available research identifies North America as a positive outlier in emissions). It would also help to pinpoint policy actions that can encourage consumers to reduce emissions.

Avoid-Shift-Improve

An easy framework for thinking about lifestyle changes, developed by Felix Creutzig and others, is Avoid-Shift-Improve. Avoidance is best understood as reducing the overall level of consumption. For example, a key “ask” of consumers is to avoid long-haul and medium-haul flights, as these have considerable carbon emissions associated with them. Smaller houses, reductions in food waste, and living car-free (thanks to the availability of car sharing through businesses like Uber) can be added to this list. On shifting, use of public transport, shifting diets to reduce beef and lamb consumption, and buying local produce are part of the answer. On improving energy efficiency, transitioning to electric cars and purchasing sustainably produced products are the main drivers.

In each of these cases, there is a public policy reason to encourage the shifts in consumer behavior, and this is most readily achieved through differentiated taxes. There is plenty of talk about taxing carbon, but a uniform tax on carbon is not an efficient nor a fair solution. Application of the Ramsey optimal tax rule would suggest that the appropriate tax rate on a good be proportional to the unit contribution of that good to carbon emissions, and inversely proportional to the elasticity of demand of the good in question. This is the set of taxes that would minimize the deadweight welfare loss stemming from taxation.

With this in mind, there are three priorities for lifestyle changes:

1. Impose a tax on the main areas where “avoidance” is the priority—air flights, beef, and lamb are huge sources of carbon. In today’s world, they should be considered a luxury and taxed accordingly. As these industries reduce emissions (for example, the addition of kelp to animal feed appears promising in reducing methane emissions), the optimal tax rate should decline.
2. Use the revenues to subsidize the choices to which consumers should shift—public transport and local food producers of vegan products.
3. Set standards and encourage business research to develop efficient appliances, most importantly electric cars, trucks, and buses.

The Ramsey rule tells economists how to set relative tax rates across goods. The absolute level of taxes depends on how much revenues need to be raised. Similarly, the level of the Ramsey carbon taxes described above depends on the global carbon budget that must be respected. In practical terms, much of this depends on global population. I have written before on how the best investment in reducing carbon emissions is investing in girls’ secondary education in high-fertility countries. This remains true—by a large margin. It is disappointing that at COP26, even in the special session on climate change and health, there was no mention of education.

COP26 missed an important opportunity to highlight the role that lifestyle changes can bring about, and to prepare the ground for policies that will surely need to be implemented to make these lifestyle choices acceptable to the population. People accepted taxes on cigarettes as a tool to reduce smoking and increase life expectancy. They need similar efforts to understand why flying and eating meat deserve similar treatment. What’s more, they need to understand that the demographic impact on each of us of lower population growth in poor countries is large. That’s why combining climate finance and development finance makes so much sense.

So let’s not forget about what each of us can do to consume in more responsible ways. Let’s also understand the big picture. Reducing food waste is good, but small compared to avoiding a single flight. Reducing meat consumption is as important as giving up your car. At the time of COP27, when more ambitious targets and actions are expected, we should look for proposals to encourage low-carbon lifestyle choices. And we should redouble efforts to accelerate girls’ education.

      

The time has come for international agencies, donors, and NGOs to adapt their business cultures and operating models to a new era of international development cooperation. The converging transformational forces of the COVID-19 pandemic and the movement to “decolonize development” have created new options for development partnerships to move away from the traditional use of expatriates in management and technical roles.

The first transformational force was the COVID-19 pandemic. By the middle of March 2020, international travel had shut down and organizations were scrambling to shift to remote work. While the capacity for virtual collaboration and remote work via video conferencing had been available for at least 10 years and there was some movement in this direction, the forced shift to a remote posture shattered cultural and perceived operational barriers to working remotely overnight. By May 2020, surprisingly few programs that had relied on external oversight and on-the-ground technical assistance had closed, and development organizations of all types (donors, multilateral development banks, NGOs) were expressing pride in how quickly they had adapted to collaborating and providing management and technical services virtually.

The converging transformational forces of the COVID-19 pandemic and the movement to “decolonize development” have created new options for development partnerships to move away from the traditional use of expatriates in management and technical roles.

The second transformational force that emerged in 2020 was the rise of a powerful movement around the need to “decolonize development.” While the decolonization critique is not new and echoes the 1970s outcry against “neo-colonialism,” the current manifestation has fired the passions of a new generation of development professionals caught up in the racial and social reckonings shaking the U.S. and Western Europe. In the context of international development, this translates into demands to address the power imbalances between donor nation and developing country organizations and professionals. In practical terms, it means greater transparency and local participation in how development assistance is programmed, greater local control over spending decisions, more use of local institutions and expertise, and achieving pay equity among international and national employees.

These two convergent forces have fundamentally changed the operating environment for international development organizations and cleared the path for more sustainable and cost-effective approaches to collaboration between international and local organizations, including the home office and country operations of large international organizations.

A new model for development cooperation

New approaches to project oversight and technical assistance are already embedding themselves in development organizations’ operations, but it would be useful for the development community to articulate these emerging practices as a preferred operating model guided by two simple principles:

First, whenever possible resident management and technical staff should be hired locally. The advantages here are well known: Local professionals possess the language and cultural skills and community networks essential to effective development work. Moreover, with the notable exception of most conflict-affected states, the excuse that local professionals are unavailable is no longer valid. A common and legitimate critique of international organizations is that the higher salaries and benefits they offer to expatriates are inherently inequitable, reinforce old colonial power imbalances, and can distort national labor markets. This first principle resolves these problems by eliminating differential treatment and employing all resident staff on a single set of terms and conditions of service that conform to the local labor market.

The second principle is that when external management and technical expertise are required, they should be provided virtually to the greatest extent possible. This recognizes that international collaboration is critical to addressing today’s development challenges, that complex endeavors frequently require highly specialized experience and skills, and that there is value in being part of larger international professional networks and initiatives.

Here it is worth digressing to address a weakness in the current decolonization narrative: the implication that donors need only provide the financing and leave recipient countries to handle the rest. This ignores the value of international collaboration in promoting innovation and technology transfer and in building capacity. All countries—rich and poor—are better off when they have economic, scientific, social, and institutional linkages to their neighbors and the larger international community. Going it alone is neither politically feasible nor practically desirable.

That said, the old argument that international organizations require expatriate on-the-ground presence to achieve results no longer holds water, thanks to the experience of working virtually over the last two years. However, embracing new practices does require changing organizational culture, as the allure of international travel and in-person collaboration are major motivators for many development professionals. While the approach proposed here does not eliminate all travel—certainly there is value in some in-person interaction (for example, to establish relationships, understand context, and conduct some types of research), the amount of international travel will greatly diminish. On the plus side, this will help international organizations reduce their carbon footprint, but as the new model takes hold, expect many U.S. and European development professionals to retire or change careers and many developing country counterparts to lament the reduction in opportunities to travel abroad.

So, what does this new model look like in practice?

It’s simple. All in-country positions are treated as local national positions with compensation and conditions of service geared to the local labor market. This doesn’t prevent an expatriate from competing for a position, but it removes the financial incentives to use expatriates.

Concurrently, short-term expatriate assignments, with rare exceptions, are virtual. This is far more economical, more environmentally friendly, and tilts power relations in favor of finding local employees while avoiding distortions to local labor markets.

Forty-five years ago, Ross Coggins’ satirical poem, “The Development Set,” summed up the contradictions and weaknesses inherent in the classic international development model. In 2020, the forced shift to remote work and a renewed and urgent concern about inequalities embedded in the development community’s traditional operating arrangements have created the conditions to replace “development tourism” with a new approach to development cooperation. Now, it is incumbent upon donors, multilateral institutions, and NGOs to take action.