7 The Future of Global Infrastructure This volume has been premised on the assumption that plays a vital role in a country s development by underpinning economic growth and enabling human development. Thus, it was surprising to us that few efforts have sought to explore long-term futures for in a comprehensive fashion that is, looking at an extensive range of s across a wide variety of countries over a long time horizon and linking to other key human development systems. In this volume, we make contributions on two main fronts. First, we add to the forecasting toolkit through building and making available an empirically based, more comprehensive and accessible computer modeling system for thinking about alternative futures. Second, we use the tool to explore where the world seems to be headed in terms of development; how this compares to existing targets; how the pursuit of these targets might interact, positively or negatively, with broader human development; and what might be alternative targets. This chapter reflects on both of these contributions. Expanding the Capability for Forecasting Infrastructure Our goal for Building Global Infrastructure was to create a dynamic, integrated approach for forecasting stocks, access, and spending for 183 countries around the world and for a long time horizon. Furthermore, we wanted to embed this structure into the International Futures (IFs) system, with its broad representation of multiple domains, such as economics, demographics, energy, food and agriculture, governance (including revenue generation and public spending), and the environment. This would allow us to incorporate driving variables not used in previous forecasts (e.g., those related to income inequality, geography, and governance); to take into account fiscal constraints; and to explore the wider socio-ecological impacts of different futures. This volume makes contributions on two fronts: (1) improving the toolkit for forecasting, and (2) exploring the future of and its broader implications. The Future of Global Infrastructure 171
IFs provides a dynamic, integrated approach to forecasting key stocks, access to them, and the associated spending for 183 countries out to the year 2100. In developing this forecasting approach and embedding it into IFs, we ran into a number of related challenges. These included a lack of comprehensive and reliable historical data and somewhat limited, and at times contradictory, empirical information on the drivers of development and the effect of on other socioecological systems. Unlike other human development issue areas, including population, education, health, agriculture, and energy, no lead international organization deals with as a whole, nor does the field have a standard set of terms and measurements. Thus, data have long been scattered across a number of sources, resulting in data that are difficult to obtain and reconcile across time, geography, and types. This is true for physical data, but even more so for financial data. These data problems exacerbated the difficulties for us, and many others, not just in forecasting development, but also in estimating the impacts that can have on human development. Thus, it is not surprising that most previous forecasting efforts have narrowed their focus to select countries, considered a single or small subset of types, maintained a relatively near-term focus, and/or paid attention primarily to needed or targeted levels of and their related funding requirements. Further, the studies that did look at costs did not explicitly address the financial and other constraints that must be considered in forecasting actual levels of (that is, they did not consider the revenuegenerating capacity of governments or the competition for funds from health, education, and other categories of public spending). Finally, previous studies very seldom considered the impacts of alternative futures on broader human development systems (such as on economic growth and on aspects of demographics, including health), much less closed the loop by exploring how the resultant alternative futures for those other systems in turn further affect development. An additional challenge we faced is that technologies continue to change (this has been seen most recently in the very rapid advances in information and communication technologies). Although our model is more structural than extrapolative, it does begin with the current state of global ; further, it links near-term development patterns to the general trends of recent years, and its structural elements have been designed, in large degree, based on historical patterns. Thus, even though we know there will continue to be new technologies and new forms of, we are not able to forecast with any confidence what they will be, what they will cost, or what patterns of adoption will characterize them. These challenges mean that any effort to understand the future of, including our own, will be limited in important ways. Nonetheless, we have made what we believe is a significant contribution in enhancing the toolkit for forecasting. A few of the key characteristics of our model are that it: provides annual forecasts for 183 countries out to the year 2100 (although the time horizon for the results presented in this volume extends only to 2060); includes a wider array of forms than most previous studies, adding multiple information and communication technologies (mobile telephones, mobile broadband, and fixed broadband) as well as expanding more standard types such as road transportation (to include unpaved roads and rural road access) and water and sanitation (to include wastewater and irrigation); explicitly considers financial constraints on development through the estimation of a supply-demand balance imposed in our forecasting of actual levels of and associated spending; uses stocks and access as direct inputs into the determination of a number of economic and health outcomes; feeds back the direct and indirect effects of development to affect the drivers of development in future years; enables users to create their own scenarios by defining alternative targets for attainment and changing a number of underlying parameters; 172 Patterns of Potential Human Progress Volume 4: Building Global Infrastructure
is freely available to run or download from our website (www.ifs.du.edu); this includes access to what we believe is the most comprehensive database of stocks and access currently available. Chapter 4 provides greater detail on many of these aspects of the model, as does the help system included with the IFs system. Understanding the Future of Infrastructure In the past, as new technologies were invented, today s high-income countries implemented in generally sequential patterns and over relatively long periods of time the networks they now enjoy. Today, however, developing countries are faced with an array of existing technologies and populaces who desire access to all of them simultaneously. How do people s desires for all basic s compare with the reality in today s developing countries? In 2010, the with the widest distribution was that for providing access to safe water. In that year, 88 percent of the world s population had either household connections or access to some other improved source of water. Still, 789 million people had no such access, including 333 million in sub-saharan Africa (66 million in Nigeria alone) and 158 million in South Asia (94 million in India alone). Access to electricity and to mobile telephones appears to be the next most pervasive, with global access around 74 and 77 percent, respectively. For mobile phones, this might be somewhat misleading, however, as the measure, subscriptions per 100 persons, can reflect multiple subscriptions held by a single person. Still, the 2010 global figure of nearly 78 mobile phone subscriptions per 100 persons points out the rapid penetration of this technology in recent years. We are seeing a similar rapid increase in mobile broadband subscriptions which, in fact, already outpace those for fixed broadband (the greater subscription rate for mobile broadband is related to the greater physical investment required for the fixed technologies, the availability of mobile services in many areas where fixed services are unavailable, and the greater convenience of the mobile technologies even when both fixed and mobile services are available). Access to improved sanitation and rural access to allseason roads lag behind access to other forms of basic. The variation in current access patterns across income levels and geographic regions is often very large. This is particularly pronounced for electricity and improved sanitation, where the access rates in 2010 for sub-saharan Africa were only 29 and 30 percent, respectively, compared to 98 percent or more in high-income countries. The pattern of access also varies for individual countries within income groups and regions, but in general, low- and lowermiddle-income countries have greater access to improved water than to sanitation or electricity. There is further general consistency across countries and regions in that those that rank lower in access to one form of also tend to rank lower in other forms. However, the spread of new technologies is facilitating more rapid advance in today s developing countries than high-income countries experienced in earlier periods. The clearest example of this is the expansion of mobile information and communication technologies, which is allowing many countries to forego altogether the widespread development of fixed telephony. Unfortunately, for many countries progress in development has occurred in fits and starts, reflecting inherent challenges: high upfront costs, long time-frames required for new construction, and the important roles that governments (sometimes poorly functioning or corrupt governments) need to play in development and provision. In addition, efforts to increase private participation in development have been uneven (as have been the results of such efforts). Further, in some countries, for example, Haiti and Iraq, natural disasters and conflict have taken a severe toll. Of course, many of these factors also affect high-income countries, but the greater challenges for these countries are related to maintaining existing and keeping pace with new technology developments rather than building out traditional. Despite these problems, using our Base Case, we forecast that great progress in access rates is likely around the world by 2060 (see again We believe we have provided some valuable insights into the possibilities and implications of alternative futures for human well-being. In the future, developing countries will not follow the same path of development that today s rich countries once did. For many countries, progress in improving access to has occurred in fits and starts, reflecting a number of inherent challenges. The Future of Global Infrastructure 173
Despite the challenges, great progress in provision is likely in most countries over the next half-century. The pace of development in our Base Case is not fast enough for many countries to achieve universal access to most basic forms by 2030. Table 5.3). Information and communication technologies (ICT), already an incredible success story, will be all but universal well before 2060. We forecast that, with the exception of sub-saharan Africa, most developing regions will have access rates to improved water and electricity at levels that approach or exceed those in high-income countries today. Even sub-saharan Africa will reach 81 percent access to electricity and 91 percent access to improved water. Access to improved sanitation will approach 90 percent globally. Only South Asia (87 percent) and sub-saharan Africa (77 percent), which start from very low levels of current access, will have regional levels below the global average. Overall, road networks will grow substantially in all regions; however, although rural access to all-season roads will increase, it will not grow much beyond 80 percent in a number of regions and not beyond 60 percent in low-income countries. These rural access forecasts are related to assumptions in IFs which reflect historical data that indicate rapidly increasing marginal costs to increase rural road access above 80 percent. Overall, therefore, the IFs Base Case projects a generally optimistic scenario that also includes positive or virtuous feedbacks across s advance and broader human development. 1 For example, in 2060, the number of people without improved water and sanitation will be reduced by more than 50 percent compared to 2010, even as we forecast that the global population will increase by 40 percent over the same period. Still, large gaps and shortfalls will remain. Our forecasts suggest that 1.1 billion persons will not have improved sanitation in 2060, and 240 million will lack access even to an improved water source. Furthermore, those without access in 2060 are likely to be even more concentrated in sub-saharan Africa than they are today, with the continent being home to half of those without improved sanitation and more than 85 percent of those without access to improved sources of water. The pace of advance in our Base Case is also not fast enough for 82 percent of countries worldwide, and 99 percent of developing countries, to meet the combined goals put forth across the separate studies that call for universal access to various forms of basic by 2030 (see again Chapter 1 and Box 5.3). This failure is a reflection of a number of factors, including population growth, competing public priorities, and perhaps most important, poor capacity for revenue generation (particularly in the least developed countries). As a general rule, both low-income and middle-income governments spend a significantly higher percentage of their GDP on core than do those with high incomes in fact, approximately four times as high (about 4 percent compared to a bit less than 1 percent). 2 While low- and lower-middleincome countries, in particular, are likely to maintain or even increase that rate over the next two decades, our general expectation is for decreasing percentages to go toward core in all countries because other categories of spending, notably education and health, tend to take higher percentages of GDP as countries income levels rise and overall government spending as a percentage of GDP remains relatively constant or declines. Even if we push governments to achieve universal access for core, as we do in our Universal Targets Pursuit scenario, 78 countries (43 percent of all countries and 54 percent of developing countries) are unable to do so by the 2030 target date if they must rely only on their own resources (see again Table 6.3). By 2060, 84 percent of the world s countries, and 78 percent of developing countries, are able to do so (compared with only 32 percent of all countries and 13 percent of developing countries in the Base Case). The vast majority of countries that are not able to achieve the universal targets are low-income countries, primarily in sub-saharan Africa. The principal reasons they are unable to meet the targets are that they have further to go and they do not have the financial resources to do so, in spite of diverting significant amounts of public funding from other important spending categories (like health and education, see again Table 6.4). Furthermore, because most of the other public spending categories also contribute to economic and human development, the negatives caused by the diversion of funds to may outweigh the benefits that come with the increased. To explore this, we looked at the cumulative discounted values of 174 Patterns of Potential Human Progress Volume 4: Building Global Infrastructure
GDP per capita and the Human Development Index (HDI), and defined the concept of a payback horizon as the time horizon beyond which the cumulative discounted values of GDP per capita or HDI in the Universal Targets Pursuit scenario are consistently higher than those in the Base Case. When using the HDI as the underlying metric of performance, we find that approximately 60 percent of all countries have a payback horizon prior to 2030 and about 75 percent have a payback horizon earlier than 2060 (see again Figure 6.8). However, for low-income countries, these percentages fall to just under 15 percent and just above 20 percent for horizons of 2030 and 2060, respectively. When using GDP per capita as the metric of performance, however, the percentage of countries reaching payback horizons by 2030 and 2060 increases, and this is especially true for low-income countries. For the world as a whole, 70 percent of all countries have a payback horizon by 2030 and 82 percent by 2060. The corresponding values for low-income countries, again using the GDP per capita metric, are 39 percent and 56 percent, respectively. These results imply that achieving universal access by 2030 is probably neither politically acceptable nor economically and socially desirable for a significant number of countries in the absence substantial outside assistance. We estimate that the level of assistance required is on the order of 3.6 trillion dollars. Compared to the Base Case, where official development assistance (ODA) for all development efforts reaches 1.8 trillion dollars by 2030, ODA would need to more than double over the period 2010 2030 and be committed totally to. Alternatively, foreign direct investment (FDI) would need to increase by more than 10 percent over the same period (see again Chapter 6). 3 As alternatives, we explored a set of what we hope are still aggressive but more reasonable targets by: delaying the target dates; prioritizing specific types of ; and adjusting the target levels in a way that takes into account the very different starting points of countries both across and within levels of development. In all, we looked at ten different scenarios (the Base Case, the Universal Targets Pursuit scenario, and eight modified target scenarios) and their implications for human development, as measured by both cumulative discounted GDP per capita and the HDI. Our results suggest that the most effective target set varies by country and changes depending on the length of the policy horizon (see again Chapter 6 for more details). We do find some general patterns based on a country s starting income level. Using the HDI metric, the scenario targeting only improved water and sanitation for universal access provides the greatest returns for an increasing number of low-income countries as the policy horizon increases. (When the policy horizon is set to 2060, almost three quarters of all low-income countries benefit the most from this targeted intervention.) The ICT only scenario provides the second-largest return for low-income countries when shorter policy horizons are considered. For lower-middle-income countries, the targeted Water and Sanitation Only and ICT Only scenarios provide the biggest benefit with shorter policy horizons, but are replaced by the Universal Targets Pursuit scenario as the policy horizon extends to 2060. A majority of the upper-middle- and high-income countries see the most benefit from the pursuit of a full suite of targets, usually either the High Performance scenario (where countries attempt to achieve access rates equal to the best performing country in their income group) or the Universal Targets Pursuit scenario (see again Figure 6.13 for a fuller picture). Based on these findings, we are confident that some acceleration of development beyond that seen in the Base Case is warranted in almost all countries, although the optimal rate and focus of this acceleration will differ across countries. 4 We can be certain, however, that currently developing countries will not follow the same general path of development that today s rich countries once did. On the one hand, developing countries are under pressure for rapid initial build-out of multiple traditional forms simultaneously, a formidable task. On the other hand, new technologies especially, but not only, mobile information and communication technologies will allow them to leapfrog over some of the intermediate steps that have characterized the past transitions of today s rich countries. Pursuing universal access by 2030 in the absence of substantial international assistance is probably neither politically acceptable nor economically and socially desirable for a significant number of countries. Some acceleration of development beyond that seen in the Base Case is warranted in almost all countries, but the optimal rate and focus of this acceleration differs by country. The Future of Global Infrastructure 175
The Future of Infrastructure Forecasting In this volume, we have painted a generally bright future for development based on current expectations. We have also noted that it could be brighter, if supported by significant and careful policy choices, both within and across countries. Creating these brighter futures will likely require not only greater international assistance and participation of the private sector in provision, but also wise choices by domestic governments in terms of how best to accelerate development. Furthermore, the path to these futures will likely differ between countries. Our conclusions are very much contingent on the tool we have developed to explore alternative futures. From the start, we recognized that we could not create a fully comprehensive model. Even so, we would not have undertaken such a daunting task had we not believed both that considering the long-term future of globally is important, and that we could provide some insights into possible futures and the implications of alternative futures for broader human well-being. Much more can be done to further strengthen forecasting. In particular, we would point to the need to strengthen the database of current and past stocks of, access to, and spending on ; broaden the inclusion of types (e.g., ports and airports), as well as inter-country and global linkages; expand the treatment of interactions across types of (e.g., the potential impact of ICT on energy systems via smart grids); and improve the understanding and inclusion of important forward linkages from (e.g., to education and the environment). Finally, the intersection of hard with soft (e.g., knowledge systems and governance) remains an important area for exploration in advancing society s understanding of and futures. 5 We hope that most of our insights will hold up in the face of future efforts to advance the collective enterprise of building global. The end tables that accompany this volume provide the Base Case results for an array of and other human development indicators for 183 countries, as well as regional groupings of countries, for the years 2010, 2035, and 2060. Detailed results for other forecast years and the other scenarios discussed in this volume are packaged with the model itself. We encourage interested readers to learn from our endeavors, to take to heart our key messages, and to explore their own ideas using IFs, which is available at www.ifs.du.edu. 1 There are, however, a number of potential storm clouds on the global forecast horizon relating to: aging populations and the fiscal pressures associated with them; forthcoming peaking of conventional oil and gas production and the need to develop alternative, sustainable sources of energy; growing pressure on fresh water supplies; and accelerating climate change. We have therefore also developed and explored elsewhere a Global Challenges scenario that heightens all of these pressures relative to the Base Case (see B. Hughes, Irfan, et al. 2011). We have not explored the implications of such a scenario in this volume. 2 These expenditure estimates include only public spending on the forms of core explicitly included in Ifs. The numbers presented here do not include public spending on other forms of or private spending on any type of. 3 These figures are based on IFs Base Case estimates of ODA, FDI, and the funds needed to achieve the universal targets while avoiding any diversion of public spending on other categories. 4 Our forecasts for individual countries or regions should not be taken as specific recommendations, but rather used to provide guidance and methodological suggestions for more detailed analysis. 5 Our first chapter noted that the knowledge systems of countries, including education systems (perhaps especially at the tertiary level) and also R&D systems and government support for them, are also a form of, somewhat softer than the traditional physical forms, but also involving many physical (e.g., classrooms and research facilities) as well as institutional and cultural components. One could well argue that knowledge systems, in interaction with ICT, may be the key transformative elements of development over the next 50 years. In fact, within IFs we have begun to develop forecasting around knowledge systems, and we wish to build and use this more fully in future analysis. 176 Patterns of Potential Human Progress Volume 4: Building Global Infrastructure