14 October 2018 (Desdemona Despair) – On Monday, the International Monetary Fund (IMF) released its semi-annual World Economic Outlook report for 2018. Twice a year, Des looks forward to browsing this hefty document for the latest financial and energy data, and this year’s report, Challenges to Steady Growth, has lots of good stuff. Here are a few graphs that caught my eye.—Geopolitical Risk Index, 2010-2018Geopolitical Risk Index, 2010-2018. Geopolitical risks continue to trend upward. Data: Caldara and Iacoviello, 2018. Note: ISIS = Islamic State. Graphic: IMF

Page 23: A range of other factors continues to influence the medium-term outlook in various regions. Geopolitical risks (Figure 1.19) and domestic strife are weighing on the outlook in several economies, especially in the Middle East and sub-Saharan Africa. Box 1.5 documents the depth of macroeconomic distress in several countries (such as Libya, Venezuela, and Yemen) and compares it to other cases of large GDP collapses in recent history. While the baseline forecast assumes a gradual easing of existing strains, an intensification of conflicts in the Middle East and Africa not only would have large negative domestic repercussions (Box 1.1 of the April 2017 WEO), but could trigger a rise in migrant flows into Europe, potentially deepening political divisions. In several systemically important economies, declining trust in national and regional institutions may increase the appeal of politically popular but unsustainable policy measures, which could harm confidence, threaten medium-term sustainability, and, in the case of Europe, undermine regional cohesion. Furthermore, many countries remain vulnerable to the economic and humanitarian costs of extreme weather events and other natural disasters, with potentially significant cross-border ramifications through migration flows.

G20 Advanced Economies Projected Public Debt Change, 2017–22 (above) and G20 Emerging Market and Developing Economies Projected Public Debt Change, 2017–22 (below)

G20 AEs Projected Public Debt Change, 2017–22 (above) and G20 EMDEs Projected Public Debt Change, 2017–22 (below). Data: IMF staff calculations. Note: AEs = advanced economies; EMDEs = emerging market and developing economies; G20 = Group of Twenty; WEO = World Economic Outlook. Graphic: IMF

Page 25: Above-trend growth in many advanced economies offers a chance to build fiscal buffers and prepare for the next downturn. Figure 1.21 highlights that, while public debt is projected to decline in many of the largest advanced economies over the next five years, projected changes in public debt are uncorrelated with initial debt levels.12 Procyclical fiscal stimulus should be avoided and rolled back (for example, in the United States), while further steps should be taken by countries with fiscal space and excess external surpluses to boost domestic growth potential and address global imbalances (for example, in Germany).In cases where fiscal consolidation is appropriate, the pace of fiscal tightening should depend on economic conditions and avoid exerting sharp drags on demand, and efforts should be made to reorient the composition of spending and revenues to enhance inclusiveness and protect vulnerable people. Fiscal spending should prioritize areas that can support growth, such as investing in physical and digital infrastructure, boosting labor force participation where aging threatens future labor supply, and enhancing workforce skills.In the United States, the tax overhaul and higher spending will widen the fiscal deficit, which was already set to deteriorate over the long term because of aging-related spending. Against the backdrop of record low unemployment rates, the deficit expansion is providing a short-term boost to activity in the United States and many of its trading partners, but at the cost of elevated risks to the U.S. and global economies. The larger deficit not only will leave fewer budget resources to invest in supply-side reforms, but will add to an already-unsustainable public debt and contribute to a rise in global imbalances. With the U.S. economy already operating above potential, expansionary fiscal policy could lead to an inflation surprise, which may trigger a faster-than-currently anticipated rise in U.S. interest rates, a tightening of global financial conditions, and further US dollar appreciation, with potentially negative spillovers for the global economy. The preferred policy course would be to increase the revenue-to-GDP ratio through greater reliance on indirect taxes.

Ongoing episodes of large declines in GDP per CapitaOngoing episodes of large declines in GDP per capita, 1965-2017. The graph shows ongoing episodes of large declines (20 percent or moe) and the share of affected countries (right scale). Graphic: IMF

Page 46: A number of countries, including Greece, have suffered very large declines in GDP per capita in the aftermath of the global financial crisis. In some countries affected by conflict, such as Libya, South Sudan, Syria, and Yemen, ongoing declines in GDP per capita have been staggering.1 In Venezuela, GDP per capita is estimated to have declined by more than 35 percent over 2013–17 and is projected to decline by close to 60 percent between 2013 and 2023. Are these episodes rare occurrences? To address this question, this box documents the frequency and characteristics of large declines in GDP per capita over the past 50 years. It shows that such episodes are unfortunately not rare. They tend to be protracted and originate from a variety of sources, and the post-trough recovery, in many cases, is insufficient to even restore the starting level of GDP per capita. The chosen threshold (a decline in GDP per capita of at least 20 percent from peak to trough) is designed to isolate extreme episodes, typically occurring over several years, rather than more frequent cases of macroeconomic distress (caused, for example, by a financial or exchange rate crisis). There is a vast literature on the macroeconomic implications of different types of crises (financial, external, currency, banking, fiscal). While these crises are typically associated with severe macroeconomic distress, such distress rarely causes a decline in the level of GDP exceeding 20 percent. The literature on large GDP declines is relatively small. An important study in this respect is by Becker and Mauro (2006), who examine output drops in a large panel of countries and systematically relate them to a variety of shocks (terms-of-trade declines, financial shocks, wars, and so on). A related literature looks at large declines in GDP and consumption (“disasters”) with the objective of calibrating the impact of these rare events on financial market variables such as equity premiums (see, for instance, Barro and Ursua 2008; Barro and Jin 2011; Nakamura and others 2013). These studies typically rely on long time series data (stretching to the early 19th century) for advanced economies and a few emerging markets.2 There are four main causes, often intertwined, of GDP declines in the sample under consideration. These include strife (war, civil war, armed rebellion), commodity shocks,3 crises (including banking crises, external crises, and so on), and the transition from a centrally planned to a market economy. Misguided macroeconomic policies during the episodes play a role in a number of cases as well, often interacting with other factors. Prime examples are cases of hyperinflation, including the ongoing case of Venezuela. Declines attributable to other causes (for example, natural disasters) are much less frequent—the one example in the sample is the 2015 Ebola epidemic in Sierra Leone.

Stylized Facts on Sharp GDP Declines

The 133 episodes of large GDP per capita declines identified in the period 1960–2017 are listed in Table 1.5.1.4 They affect 92 countries (a number of them repeatedly).5 Figure 1.5.1 depicts the number of ongoing episodes of sharp declines in GDP per capita by year, as well as the share of countries affected (in relation to the total number with available data). The figure indicates that the lion’s share of episodes took place during the 1980s, following the global economic downturn and the 1982 debt crisis. The number of episodes declined in the late 1980s but rose again in the early 1990s because of the GDP declines associated with the transition to a market economy in countries of the former Soviet Union and in central and eastern Europe. The number of ongoing episodes has since declined sharply, despite some increase associated with the global financial crisis and its aftermath. Episodes associated with war are the most frequent, followed by commodity shocks, crises, and transition.

The Aftermath of GDP Declines

The focus now turns to the speed at which GDP per capita rebounds after these sharp declines. For that purpose, the analysis considers both the growth rate in the five years following a trough as well as the length of time it takes for countries to return to their pre-decline levels of GDP, and explores whether these variables are correlated with basic characteristics of the episodes: the initial level of development, the size of the country, the extent of the GDP decline, and the duration of the episode. Constructing these post-decline variables reveals a striking stylized fact: out of the 92 countries experiencing a sharp decline in GDP per capita in the sample, 45 had GDP per capita in 2017 still below its pre-decline level.These countries account for over 5 percent of global GDP at purchasing power parity in 2017, and about 7½ percent of world population. They are predominantly small. Exceptions include Iran, Ukraine, Venezuela, and some economies in the Gulf Cooperation Council with high GDP per capita that have experienced very rapid population growth, including because of immigration (Kuwait, Qatar, Saudi Arabia, United Arab Emirates). Excluding these four countries, those countries still below their past peak in GDP per capita account for about 3 percent of global GDP. […]A surprising result is that the post-decline growth rate is uncorrelated with the extent of the previous change in GDP per capita, holding constant the length of the episode. In other words, deeper downturns are not followed by sharper recoveries. However, the post-decline growth rate is strongly negatively correlated with the length of that decline. The regressions also suggest that, on average, recoveries tend to be weaker in smaller countries, consistent with the evidence on challenges to economic performance in small states.

Projected Change in the Working-Age Population Change in the Working-Age Population (15–64) Relative to 2015 Levels (Millions) in the advanced economies, low-income developing countries, China, and the rest of the world, projected to 2050. By 2035 the number of people in low-income countries reaching working age (15–64) will exceed that of the rest of the world combined. Data: UN (2017); and IMF staff calculation. Graphic: IMF

Page 30: Under current policies in many low-income countries, per capita income growth is projected to remain sluggish and below past averages. Many low-income countries are also facing pressure to accommodate a rapid increase in the working-age population. It is estimated that by 2035, the number of people in low-income countries reaching working age (15–64) will exceed that of the rest of the world combined (Figure 1.22). Creating enough jobs to absorb the new entrants will be vital for welfare and social and political stability. In this regard, economic diversification into labor-intensive activities outside agriculture, and away from excessive dependence on commodities for resource-intensive exporters, is critical. While the manufacturing sector has traditionally served as a source of well-paying jobs for low- to middle-skilled workers in developing economies, market services sectors such as retail, transport, telecommunications, and financial and business services can be viable alternatives (Chapter 3 of the April 2018 WEO). Facilitating private sector development—including by strengthening investor rights and the rule of law, reducing the cost of doing business, and enhancing infrastructure and openness to trade—would help strengthen investment and growth. Improving education standards will be essential to ensure that the growing pool of workers has the necessary skills. Achieving robust growth will also require enhancing the macroeconomic resilience of low-income countries, including against climate change. Stronger buffers and sound macroeconomic policy frameworks, alongside policies and institutions that make it easier for labor and capital to move across economic sectors and geographic regions, are essential to that end. To reduce adverse consequences from climate change, countries could also invest in specific adaptation strategies that reduce exposure and vulnerability to weather shocks, such as climate-smart infrastructure, the adoption of appropriate technologies and regulations, and putting in place well-targeted social safety nets that can promptly deliver support (Chapter 3 of the October 2017 WEO).—Energy Demand and GDP per Capita

Energy demand and GDP per capita for six nations: China, India, Japan, Russia, France, and the U.S. Data: International Energy Agency; World Bank, World Development Indicators database; and IMF staff calculations. Note: Adjusted fitted values show the S-shaped energy-income relation (constructed using the cubic polynomial) while energy demand per capita is adjusted for estimated time fixed effects. Estimates are from the baseline specification. Graphic: IMF

Page 56: The estimated S-shaped energy-income relationship (Figure 1.SF.4) not only predicts energy demand growth to be highest in emerging markets but also captures the behavior of energy demand at low-income levels. Typically, in most low-income countries, energy consumption initially declines in response to income growth probably as the result of graduation from biomass (solid biofuels excluding charcoal)—an inefficient source of energy. Biomass, in fact, is an inferior good, implying that households reduce its use as income grows. The share of biomass in total primary energy supply of the country tends to decline as income grows (Figure 1.SF.5).
In conclusion, the evidence suggests that the relationship between energy demand and income follows an S-shaped curve, with an initial decline of energy demand at low levels of income followed by stages of acceleration and then saturation at middle- and high-income levels, respectively. Thus, the main driver of future energy demand hinges on the dynamics of middle-income countries. In fact, even though some advanced economies may have already reached saturation in energy demand, estimates suggest that global saturation is still far into the future. However, total energy is not all that matters. The same level of energy consumption can be the result of varying mixes of primary energy sources, which is the topic of the next section.

Primary Energy Source SharesPrimary Energy Source Shares (Percent), 1971-2015. Data: International Energy Agency; and IMF staff calculations. Note: Sample is International Energy Agency world aggregate; grey shaded area = high and volatile oil prices; nonshaded area = low and stable oil prices. Graphic: IMF

Page 56: The optimal energy mix in each country is the result of relative resource abundance, technology, and social preferences. The local relative abundance or availability of an energy source determines its local costs, while the efficiency of use in production determines its desirability (that is, its marginal benefit).7 These two factors combined help determine the relative price of an energy source. Technical substitutability across resources then determines the impact of changes in efficiency of use or relative prices on the energy mix. For example, the relative importance of oil as a primary energy source has substantially declined over time as other energy sources became cheaper (such as coal and nuclear in the early part of the sample) or more desirable to use (such as natural gas and, more recently, renewables). The link between high and volatile crude oil prices and the decline in the oil share is indeed noticeable (Figure 1.SF.6).8 Over the long term, however, efficiency is also determined by capital investment, which allows the potential of an energy source (for example, investment in solar power or natural gas infrastructure) to be better exploited. This generates a relationship between the energy mix and the stage of development (see Online Annex 1.SF.1 for further details). At medium- and low-income levels, the semi-elasticity of the oil share to income is positive as the transport sector expands (for example, car and truck ownership increases), but it turns negative at higher income levels when the stock of motor vehicles plateaus, fuel efficiency reduces gasoline consumption, and cleaner natural gas is preferred in heating and power generation. Regressions, indeed, suggest that peak oil demand may have already been reached for some advanced economies, given that their oil share declines while energy demand is close to saturation (see Online Annex 1.SF.1). In contrast, the share of natural gas seems mostly independent of income.The relationship between income and the share of coal is weak because higher incomes are associated with cleaner energy sources but also with higher electrification rates (the main driver of coal consumption). At medium incomes, however, coal has proved to be a cheap and abundant energy source able to satisfy a quickly growing demand for electricity, especially in some large, coal-abundant emerging markets, such as China and India (being a coal producer or exporter increases a country’s coal share by 10 percentage points or 18 percentage points, respectively). Hence, notwithstanding a reduction of coal intensity at the country level, the legacy of high coal usage in large and fast-growing economies led to a surprise increase in global coal intensity in the mid-2000s (Figure 1.SF.7). As China and other major emerging markets develop, however, demand for cleaner fuels is expected to increase, leading to a decline in the coal share.Although it is too early to assess the evolution of renewables, the analysis clearly points to an increase in the use of renewables in high-income countries, especially for power generation. Advanced economies, in fact, are typically highly electrified while emerging markets, as they become more urbanized and expand the electricity grid, are expected to substantially increase their electrification rate in the medium term. The projected rise of the electric car and growth in the services sector, moreover, are expected to increase the electrification rate in advanced economies, too.
The implication of higher electrification rates is important for primary energy demand. In fact, while oil saturation will probably be reached sooner than total energy saturation (as oil’s share in the mix declines), saturation for natural gas and renewables will come later. Recent sharp declines in the price of solar photovoltaic cells and government support for the development of renewables are paving the way for the rapid growth of renewables (see Box 1.SF.1). Although coal may remain attractive for some countries, local air pollution has compelled China and India, to some extent, to shift toward renewables. Thus, cost changes and environmental concerns will play a key role for the increased penetration of renewables and the saturation point for coal.
Most of the increase in energy consumption is expected to come from emerging markets whose energy demand is approximately at its peak income elasticity, which is about one. In contrast, that elasticity is close to zero for advanced economies, suggesting that their contribution to energy demand growth will be more modest or possibly absent. Nonetheless, emerging markets’ saturation point for energy demand is still far in the future—even assuming steady gains in energy efficiency. Saturation, however, is probably much closer for some energy sources, such as coal and oil, raising the risk of stranded assets for high-cost projects, while other sources, such as natural gas and renewables, are expected to become more important in the energy mix as electrification rates increase. Even though dynamics in energy transitions and technological innovations are hard to predict, substantial long-term investment is required to change the energy infrastructure of an economic system (for example, the life of power plants and airplanes is about 40 years). Nonetheless, climate concerns, energy policies, and market forces will be key in forging future energy markets as energy regulation and prices interact to stimulate or constrain technological innovation. It is the role of policymakers to exploit these interactions to develop ecologically sustainable economies.[cf. Standing perfectly still: no progress in replacing fossil fuels for the past twenty years –Des]

World Economic Outlook, October 2018