I wrote the short position paper below, with a bit of help from a colleague, for an academic audience, but never published it. Today I was thinking about it and realized that it synthesizes some of my thinking on computing and economic trends. For the last decade there has been an active area of computing research sometimes called ICTD (Information and Communication Technologies for Development) or Developing Regions computing research, largely focused on bringing to bear computing to solve the problems of industrializing nations. For the moment I’d like to set aside both the problems with the term “development” and the fact that much of such work is unlikely to succeed in the ways its proponents think it will (Toyama’s take on this is well worth reading, and is a good counterpoint to the much more mainstream ideas of the One Laptop Per Child project, among others). Nevertheless, some of the research that’s been done has been worthwhile in that it’s been forced to work within limits, both energy and financial, but many in this field aren’t that familiar with the broader issues of Limits to Growth. That’s where this paper begins.
A simple view of the world’s nations yields a comforting binary: developed nations, those that have industrialized, and developing nations, those aspiring to their example. It is in this context, for the most part, that work on development in general and ICTD in particular has operated, and by and large this has been a success and a boon to those in developing nations. In this paper, we complicate this view by adding a third, heretofore unexamined category—undeveloping regions—and discuss both the need to consider ICT broadly and networking in particular in this new context.
Recent studies indicate that energy costs will rise substantially in the coming years, due primarily to the increasing difficulty of locating and extracting fossil fuels [10, 17]. Rising energy costs, increasing demands , ecological limits , and economic consequences  will force a fundamental change in the world’s economies, and yet due to the scale of the challenge and the lack of preparation for a transition, the effects are expected to be significant; a 2005 study commissioned by the U.S. Department of Energy concluded :
“Virtually certain are increases in inflation and unemployment, declines in the output of goods and services, and a degradation of living standards. Without timely mitigation, the long-run impact on the developed economies will almost certainly be extremely damaging, while many developing nations will likely be even worse off.”
Thus the world is on a collision course: already-high resource consumption/environmental impacts in developed nations and increasing in developing ones collide with resource limits and broader ecological overshoot. We can loosely define undeveloping regions as largely today’s high socioeconomic regions that as a result of these challenges are likely, in some form or another, to face degradation of their social and economic systems. In this we also include a second somewhat distinct category: currently developing regions whose industrialization is likely to slow, stall, or reverse. In Section 2 we attempt to identify the contours of such regions in order to inform research directions.
Given these challenges, how might ICTD researchers respond? We are outsiders to the ICTD community, and as such this paper is both a challenge and a plea to the community: a challenge to consider how these new global challenges will affect the context in which ICTD work is done and a request for help from the ICTD community to apply the hard-won knowledge of today’s ICTD to these soon-to-be undeveloping regions that comprise both today’s developed world and the developing world. In these new environments, both the target objectives and the constraints are likely to be different from the environments seen today. For example, research objectives in undeveloping regions are likely to focus lower on Maslow’s hierarchy, on more prosaic but more fundamental human needs, just as ICTD work has done in developing regions, than does today’s developed-world networking research. Similarly, the environments of undeveloping regions are likely to have different infrastructure availability and financial constraints, finance and culture, than seen anywhere today. In particular, today’s developed nations often lack indigenous cultures of self-sufficiency (in large part because they were supplanted long ago) and as a result are likely to prove to be as challenge rich environments as ICTD is today.
Thus we are both proposing a new direction for research in both the ICTD community and the broader networking community, and hoping to outline the circumstances and challenges that will help inform future designs in this space; we discuss these in detail in Section 3. In particular, we contend that the goals of Appropriate Technology are ones aligned with the needs of undeveloping regions, and that using networked systems to help decrease complexity and increase transparency of socioeconomic systems will be of particular value.
Finally, there remains a sensitive topic to discuss: why those in the ICTD community should be interested in the challenges to be faced by currently-developed regions; after all, the latter have significant resources to address their own problems. A natural response would be to ignore these challenges, expecting the fates of undeveloping regions to be separate from today’s developing regions. On the contrary, however, as these challenges progress, it is likely that the circumstances and fates of those in currently-developed and currently-developing regions will increasingly be shared, as suggested by a number of analyses [10, 13, 9, 14]. It is for this reason that we do not attempt to artificially separate the two, and consider both categories within the moniker “undeveloping regions.” Beyond this matter of shared fate, we also believe that the community has significant knowledge to contribute to a shared vision of what networking and ICT can look like in this common future.
2. Case Studies
Before we attempt to define undeveloping regions, we first look at two related questions: what is the nature of near-term global challenges, and how might those challenges manifest in terms that help us identify appropriate ICTD-like responses?
For the first, we briefly summarize previous work on the subject detailing the contours of the challenges ahead. For the second, in the absence of true ethnography detailing life in undeveloping regions (since we are, after all, discussing future circumstances whose specifics are unknowable today), we must consider a substitute. As such, we examine four recent historical instances in which crises befell nations. We do this in an attempt to glean common characteristics that might inform the challenges that lay ahead, not to predict any specific crisis or circumstance. Our examples lay somewhat along a spectrum from isolated and monolithic problems to pervasive and broad-based problems: Argentina experienced a financial crisis, Greece a financial crisis along with endemic economic problems, the United States two temporary oil crises, and the Soviet Union systemic collapse.
2.1. Background In a recent paper, we described the challenges that networking will broadly face as energy and ecological limits are reached, most likely some time this decade. Specifically, we discuss energy limits—oil production in particular—why alternatives are unlikely to serve as adequate substitutes in the required timeframe, and the likely economic consequences. As a result, developed and developing nations alike are likely to face widespread economic and social challenges over at least the span of decades. We refer interested readers to a number of works on the broader topic [12, 13, 9, 14] for additional details.
In short, the impending challenges are not just due to the limits of oil production, but the panoply of issues faced by global society today. Among these are climate change , stagnant food production per capita, and broad ecological overshoot . In addition, rapid development has put developed and developing nations on a collision course: if India and China were to maintain their current rate of growth for a little over a decade, together they would consume 100% of the crude oil available on the global market (i.e. of global net exports), even under the optimistic assumption of stable global oil production .
2.2. Argentina The case of Argentina, once seen as a development success story, is fraught with geopolitical intrigue but little actual complexity of cause or effect. Blustein provides a comprehensive analysis of the economic crises that befell Argentina, in part of its own doing, and in part due to actions by organizations such as the IMF . While the cause of the challenges faced by Argentina largely differ from those described above, the societal issues faced may be instructive. Blustein describes how the impact of the financial crisis affected all social classes, the consumer economy, and the food system. Transportation systems, such as Argentina’s rail system, suffered. Despite this, perhaps due to the relatively short duration of the worst of the crisis, public health may have remained largely unaffected.
2.3. Greece The crisis currently unfolding in Greece today has similar roots as that of Argentina a decade earlier, though it bears greater similarity to the challenges we’re likely to face. As a recent report found :
By many indicators, Greece is devolving into something unprecedented in modern Western experience. A quarter of all Greek companies have gone out of business since 2009, and half of all small businesses in the country say they are unable to meet payroll. The suicide rate increased by 40 percent in the first half of 2011. A barter economy has sprung up, as people try to work around a broken financial system.
At the same time, many Greeks are returning to agriculture, and the health system is struggling to meet even basic needs . In some ways, the challenges Greece faces are harder to overcome than those of today’s developing nations, as few alternative systems—of food, finance, or health care—exist for the populace to fall back upon. While the financial origins of this crisis are well understood, there are compounding factors of energy-driven trade deficits, and the effect of high oil prices on tourism.
2.4. United States The oil crises experienced by the United States and other industrial nations during the 1970s resulted in sharp and deep recessions, and contributed to an extended period of economic uncertainty and stagnation. While the specific causes of the crises are superficially similar to the challenges we consider here, they differ in large part in that they were caused by geopolitical factors and thus were immediate but temporary whereas geological factors are likely to be gradual but fundamental. It is for this reason that Hirsch et al. note that “past ‘energy crisis’ experience will provide relatively little guidance” . Nevertheless, they serve as an empirical reminder of the dependency of industrial economies upon energy in general and oil in particular.
2.5. Soviet Union While there are numerous historical examinations of the decline and fall of the Soviet Union, the work by Orlov on this subject is of particular note for its attempt to relate challenges the Soviet Union faced with that of impending challenges faced by today’s industrialized nations . Orlov observes that many of the contributing factors and circumstances from the Soviet Union are present in the United States as well, and thus attempts to distill the ways in which the two differ as much as they share. In particular, he notes both face(d) declining oil production, unsustainable debt, unresponsive political and socioeconomic systems, and military conflict. In the post-collapse environment, he describes widespread social dislocation, shortages of basic commodities, foregone infrastructure maintenance, and the failure of new long term plans. Despite this, he notes that because the economic system was not market-based, a number of crucial sectors of society continued to function either as before or with bearable disruptions: housing, transportation, food production, medicine, education, and energy. In some of these, the Soviet systems were already somewhat dysfunctional but as a result individuals already had coping strategies (e.g. pervasive kitchen gardens); in others, the systems were government-run and thus collapse didn’t affect them (e.g. no foreclosures due to free housing; most transportation was public and supplied by domestic energy).
In contrast, the equivalent systems in free-market economies are both highly efficient but at the same time complex and interdependent; for example, one among many proximate causes of the 2008 economic crisis was the implosion of the housing bubble, which in part was caused by high energy prices which decreased demand for exurban sprawl, a cycle which fed on itself . Thus there is some reason to believe, as Orlov contends, that many of the advantages of efficient free-market economies in times of growth can become disadvantages in times of hardship.
2.6. Undeveloping Regions One remarkable characteristic is that despite the fact that these crises had different origins, took place within different cultures and socioeconomic systems, and with different historical contexts, the effects on their populations were somewhat similar (though by no means the same). Though forecasts are sure to be wrong in their specifics (though perhaps not in their broad contours) the challenges we face are likely not only to be pervasive within nations and long-lasting, but also are likely to affect nations around the world. Thus the crises themselves are likely to interfere with efforts to mitigate them if the planning is postponed. In addition, it is our hope that in anticipating these challenges we both help mitigate them and also inform the design of better systems for not only undeveloping regions, but developing and developed regions as well.
We are not expecting a specific sharp inflection point at which a nation is no longer “developed”, nor are we expecting such a process to occur quickly. Instead, we expect a slow, grinding, and mostly transparent process of undevelopment that, while we expect it to begin this decade, will not fully play out for a few decades to come. Within this context, we also expect natural societal adjustments to respond to these challenges, responses which in some cases might improve circumstances (e.g. energy efficiency programs, relocalized health, education, and food production, etc.) and in some cases might do the opposite (e.g. last ditch efforts to produce liquid fuels from coal, tar sands, etc.).
It is unclear how long lived these challenges will be—will they last on the order of a decade or two, as in these historical examples, or will they be more persistent? Some have argued that the latter is more likely because the limits faced today are fundamental and pervasive [13, 9, 14]. Nevertheless, our goal here isn’t to adjudicate the matter; a decade or two is long enough to warrant our attention. Thus we turn our attention to three categories commonly targeted by ICTD research within which we can categorize the effects of the above historical crises.
Economics. Economic challenges are the clearest common thread in these four examples. In each, financial systems struggled as liquidity vanished and the grip of inflation took hold. As a result, consumer spending plummeted (in free-market economies) and thus both the industrial and service sectors were affected. In many ways, the economic effects appeared to be like a particularly severe recession, but with significantly longer duration.
Health. A prominent concern of the ICTD community has been public health. Peak oil, limits to growth, and climate change together are likely to present perhaps the biggest set of challenges to public health that have ever been faced in modern times , so facing these issues head on is a matter of public health in developing and undeveloping regions alike. A likely cause of problems will be the mismatch between today’s expensive and complex industrial medicine and the basic needs and financial means of the populace. A recent issue of the American Journal of Public Health featured eight papers on the special topic of Peak Petroleum and Public Health with a wide array of dire findings and possible mitigation approaches.
Transportation. As we face issues with the cost and availability of oil, and thus of transportation fuels, nations with oil-dependent transportation systems will find themselves hugely vulnerable. In addition, the expense and time commitment to build alternative systems, such as an extensive rail network, are prohibitive during times of crisis, as indicated by Argentina’s experience.
3. Challenges and Responses
Networked systems design, and systems design more broadly, is always about context: about building a system that optimizes some metrics within some design constraints. Thus far we have tried to outline the issues likely to be faced by those in undeveloping regions. Here we attempt to translate such issues into concrete research challenges—constraints within which the community will need to design within and possible responses. The constraints we identify often differ not only from today’s developed regions but from developing ones as well, though the similarities are as interesting as the differences.
Before we begin, we briefly make two disclaimers. First, it is very difficult to know that the constraints we identify are the right ones; only time will tell. We have tried our best to let the case studies and other background material inform them. Second, we are aware of the issue of generalizability—that while it is of value to deliver generalizable contributions in ICTD work, there is a danger of attempting to generalize from specific instances to circumstances that may in fact differ significantly .
3.1. A Comparison
|Category||Developed Regions||Developing Regions||Undeveloping Regions|
|Network Infrastructure||Pervasive and advanced||Spotty but advancing||Varied and aging|
|Devices||Contemporary||Lagging but advancing||Lagging and stagnant|
|Power||Stable and ubiquitous||Intermittent||Intermittent but varied|
|Finance||Stable and ubiquitous||Stable but varied||Unstable|
A key difference in this challenge versus most ICTD work is that the design is in large part for the (potentially near) future, not today. In an attempt to capture the differences between the well-understood targets of developed and developing regions, and the undeveloping regions of the future, we summarize some qualities that may be pertinent to networked systems designers in the Table. In particular, we contend that undeveloping regions are likely to exhibit the unsurprising quality of slow or absent replacement of infrastructure and devices and general degradation of existing systems. The key difference between undeveloping regions and the developing regions of today is likely to be their starting point—many undeveloping regions may begin with a vast infrastructure and installed base which can serve as a buffer for some time. At the same time, they will be faced with a pervasive dependency on those systems for the functioning of interlinked social systems. Just as there are those in developing regions who are early adopters of ICT, whose combination of local knowledge and technical knowledge can prove important to the success of an ICTD project, there are some, though not many, autochthonous technology communities in developed countries , usually with an aim of independence (and sometimes sustainability).
It is just as worthwhile to examine what assumptions may no longer hold for developing regions that transition to undevelopment. Specifically, we contend that two of the three assumptions made in one of the earliest ICTD papers by Brewer et al. will no longer be true for such regions : 1) the impact of Moore’s law is likely to slow or cease, not only due to inherent challenges of scaling today’s technology at prior rates, but also due to the increased relative cost of deploying new technology during increasingly hard economic times; and 2) favorable business environments are, in some regions, likely to be a temporary artifact of global surplus and trade, something that is widely expected to reverse [12, 13, 9].
3.2. Appropriate Technology
Because transportation costs will rise for undeveloping regions, sustainability in networking and ICT will need to become a significant focus. Sustainability in networking helps solve other societal sustainability issues, and the sustainability of networking itself will have to be prioritized. In particular, we advocate the adoption of the principles of “appropriate technology” [8, 18]; namely the design of networked systems that are a) simple, b) locally reproducible, c) composed of local materials and resources, d) easily repairable, e) affordable, and f) easily recyclable. Moreover, many of the resulting objectives are part of the networking canon: to build scalable networks that can be started and tested at a small scale with modest resources, resilient networks that remain useful under changing conditions and respond to pressures, modular networks separated into distinct elements that can be replaced at different scales and technological levels, and open networks that do not demand a certain system or set of components to function. It is however the case that in recent years ICTD research has far more effectively adhered to such maxims than the broader swath of networking research.
In each of the case studies we considered in Section 2, we found that the complexity of the socioeconomic system in question was the root of much of the hardship. That is, each crisis trigger caused a series cascading economic failures. As today much of the economic infrastructure of developed economies relies upon networks and ICT more generally, we can play a direct role in decreasing complexity, or at the very least the impact of complexity. In a sense, our task is to help build resilience in the network of interactions between computing and non-computing systems throughout an economy, just as resilient and secure networks are designed to isolate faults and attacks. Such an effort would also be in line with the principles of Appropriate Technology outlined above.
A first step is to better understand the complexity of today’s networked systems. Ratnasamy began such an investigation by defining a complexity metric with which networked systems could be evaluated ; it would be valuable to extend this notion, apply it to better understand complex networked systems, and then expand the analysis to systems built upon them.
One key way to help mitigate the ecological limits being faced is to better understand them, and the tools of networked systems—in particular sensor networks and cyber-physical systems—can be of particular value. Specifically, such systems can provide scientists, engineers, and policy-makers rich databases to enable analysis of local resources (both physical and virtual resources) and their flows. This might enable local communities to realize what substitutable resources they have locally that can replace remotely sourced inputs, and also potentially learn about the ability of their local bioregion to absorb pollution and waste flows in the form of natural sinks. This information can be valuable for energy sources as well (e.g. ICT can help a developing or undeveloping region better understand the rate at which they are depleting the forest for firewood or eroding the topsoil to grow biofuels, and can help understand how sinks like air pollution relate to it). However, such systems are challenging to build because of the scale required, the unreliability of the sources, and the way in which the system will have to operate (i.e. under challenging socioeconomic conditions), which argues for their near-term development.
3.5. Development During Undevelopment
As Burrell and Toyama discuss , it is not simply an open question of what ICT looks like in ICTD, but the notion of Development itself. In the context we discuss here, we believe that the notions of development considered most prominently by the ICTD community will naturally have to broaden. While there are researchers considering development in the context of sustainability and other constraints, development in the context of undeveloping regions facing limits to growth will have to be quite different indeed. While the thus-far unstated premise of our paper—that growth and thus development as it is currently perceived and measured is in the process of ending for many if not most regions—may seem like an ideological statement, we contend that our current growth paradigm is the same: implicitly accepted ideology. Indeed the science on the subject indicates that the current paradigm cannot continue, and not in the sense of ethics, but in the sense of hard ecological limits [13, 20, 14].
Thus with the necessary transition away from this paradigm—something that we as computer scientists will likely have little role in—we can embrace those alternative paradigms that have been developed over the last several decades that provide a more productive notion of development in a post-growth era. Specifically there exists a large body of work in the domain of Ecological Economics which we believe can and should inform such thinking, and aligns nicely with the objectives of Appropriate Technology discussed earlier.
It is our belief that ICTD research will become and needs to become more central to the mission of networking research if we are to appropriately respond to the converging issues we face. The broader networking community has yet to design systems to operate under such challenging conditions, and targeted to specific and fluctuating circumstances; on the other hand, the ICTD community deals with such challenges frequently and thus has much offer. Regardless of how the issues we explore in this paper develop, infusing the methodology and knowledge of ICTD research throughout the broader community can only yield benefits.
 Blustein, P. And the money kept rolling in (and out): Wall Street, the IMF, and the bankrupting of Argentina. Public Affairs, 2005.
 Brewer, E., Demmer, M., Du, B., Ho, M., Kam, M., Nedevschi, S., Pal, J., Patra, R., Surana, S., and Fall, K. The case for technology in developing regions. IEEE Computer 38, 6 (2005).
 Brown, J. J., and Foucher, S. Egypt, a Classic Case of Rapid Net-Export Decline and a Look at Global Net Exports. ASPO-USA (Feb. 2011).
 Burrell, J., and Toyama, K. What constitutes good ICTD research. Information Technologies and International Development 5, 3 (2009), 82–94.
 Daley, S. Fiscal Crisis Takes Toll on Health of Greeks. The New York Times (December 26, 2011).
 Daly, H. Steady-State Economics: Second Edition With New Essays. Island Press, 1991.
 Hamilton, J. Causes and Consequences of the Oil Shock of 2007–08. Brookings Papers on Economic Activity 2009, 1 (2009), 215–261.
 Hazeltine, B., and Bull, C. Appropriate technology: tools, choices and implications. Academic Press London, 1998.
 Heinberg, R. The End of Growth: Adapting to Our New Economic Reality. New Society Publishers, 2011.
 Hirsch, R., Bezdek, R., and Wendling, R. Peaking of World Oil Production: Impacts, Mitigation, & Risk Management. U.S. Department of Energy NETL (2005).
 Kelly, K. What technology wants. Viking Press, 2010.
 Kunstler, J. The Long Emergency. Grove Press, 2005.
 McKibben, B. Eaarth: Making a life on a tough new planet. Henry Holt and Company, 2010.
 Meadows, D., Randers, J., and Meadows, D. The limits to growth: the 30-year update. Chelsea Green, 2004.
 Orlov, D. Reinventing collapse: the Soviet example and American prospects. New Society Publishers, 2008.
 Ratnasamy, S. Capturing complexity in networked systems design: The case for improved metrics. In Proceedings of ACM Hotnets (2006).
 Schultz, S. Military study warns of a potentially drastic oil crisis. Der Spiegel (September 1, 2010).
 Schumacher, E. Small is beautiful: A study of economics as if people mattered. Abacus, 1974.
 Shorto, R. The Way Greeks Live Now. The New York Times (February 13, 2012).
 Wackernagel, M., Schulz, N., Deumling, D., Linares, A., Jenkins, M., Kapos, V., Monfreda, C., Loh, J., Myers, N., Norgaard, R., et al. Tracking the ecological overshoot of the human economy. Proceedings of the National Academy of Sciences 99, 14 (2002), 9266.