Abstract: The Islamic State’s failure as a state was predictable as soon as the group’s initial advances stalled. The group tried to fight a three-front war for territory—Kurds to the North, the Assad regime to the West, and Iraq to the East—without the necessary resources to do so. Early revenue estimates revealed that either its revenue-generation system was inefficient, its economy had collapsed, or both, and that conditions had steadily worsened over time. The area it controlled in late 2014 was only modestly productive before the war and its governing institutions were inimical to economic growth. These factors guaranteed a slow collapse.
Writing on terrorist groups often begins with a focus on what is unique about a particular group or particular campaign. When a new group emerges, it will typically have some organizational innovation that makes it seem fundamentally different than the groups that came before. After 9/11, many focused on the “networked” nature of the jihadist threat or the particular social dynamics of recruitment into transnational plots. When the Islamic State first emerged as a major policy issue, much of the focus was on its skilled use of social media and frequent use of bloody spectacle. While covering such innovations is a natural thing to do, the focus on novelty can obscure our ability to predict what comes next or to spot the vulnerabilities inherent in innovations.
Theory provides an antidote to this tendency. While each terrorist group is sui generis in some respect, they share commonalities as well. All, for example, face a tradeoff between maximizing security for operatives and maintaining control for leaders. And the vast majority that have fully ceased activity do so either by transitioning into political parties (43 percent) or because their key members are killed or captured (40 percent). Viewing terrorist groups in light of abstract arguments about what makes a group successful or about the inherent limits they face due to organizational dynamics common to all human enterprises can be useful for anticipating how the group in question will develop. Combined with using broader historical trends to gain perspective, bringing theory to bear helps to avoid the trap of assuming that what seems novel today is indeed game changing. Sometimes it is, but more often it is not.
To illustrate this broad point about the value of theory, this piece examine what political economy and simple comparisons would have told us in early 2015 about the Islamic State’s future.[a] The factual observation underlying the argument presented here is that since the group’s remarkable advances in 2014, it has been slowly rolled back, steadily losing territory despite the occasional short-term gain such as the seizure of Palmyra in March 2015, which the government retook a year later. While this rollback is surprising from the perspective of arguments about the Islamic State’s potential that were made in the fall of 2014, most of which focused on the group’s substantial resources, arguably much of what has transpired since could have been inferred from first principles at the time. And it is the predictability of the Islamic State’s slow failure that highlights the value of theory. In the Islamic State’s case, some comparisons using widely accessible data plus two meta-theories—one about what it takes to fight a conventional war for territory and another from political economy about the institutions required for a productive economy—would have led one to predict the group’s stagnation and decline.
The Islamic State first drew significant public attention in early 2014 as it began a stunning advance into Iraq from operating areas in Syria, an advance that would culminate in the June 2014 seizure of Mosul, Iraq’s third-largest city. The Islamic State’s initial successes were followed by an offensive against the Kurdish city of Erbil and an ethnic cleansing campaign against the Yazidi minority whose broadly televised denouement eventually drew the United States and the international community into deeper involvement in the war.
While the group’s emergence in 2014 surprised many, it had deep roots in the region. The Islamic State is the successor of al-Qa`ida in Iraq (AQI) and the Islamic State of Iraq (ISI), which the Iraqi government, the United States, and their allies have fought since at least 2004. From 2006 through 2009 ISI was routed from most of Iraq by a combination of local Sunni militias, coalition forces led by the United States, and Iraqi security forces. As described in detail elsewhere, the group withdrew into a limited terrorist campaign in northern Iraq by mid-2009 and maintained a clandestine network in the Mosul area of Ninewa province. That network conducted a sustained terrorist campaign in Baghdad and other major cities as well as a targeted assassination campaign against its former enemies in Anbar and Ninewa provinces.
As the Syrian civil war picked up in 2011 ISI sent fighters into Syria under the banner of Jabhat al-Nusra. Differences over strategy, tactics, and goals between the ISI leaders in Iraq and those guiding Jabhat al-Nusra in Syria led to a split. In April 2013 ISI began operating on its own in Syria and changed its name to the Islamic State in Iraq and al-Sham, or ISIS.[b]
Antigovernment protests in Iraq’s Anbar province from January 2013 to December 2013 were met by a repressive government response, creating a political opening for the group to return to Iraq. Beginning in late December, the Islamic State expanded back into Iraq with the support of some local political organizations—including many of the same tribal organizations that had fought against it in 2006—and quickly overwhelmed Iraqi Army outposts in the major cities of Anbar. Following the Iraq Army’s precipitous retreat from Mosul in June 2014, the group renamed itself the Islamic State.
The Value of Comparison
One way to think about the Islamic State’s administrative competence and its financial prospect is to ask a simple question. Given the reported level of revenue and the best estimates we can come up with of economic activity and population in the area that became part of the “caliphate,” just before their territorial takeover, has it done a good job of maintaining an economy, raising taxes, and attracting the population necessary to fight a war over territory? On the revenue side the U.S. Department of the Treasury estimated that in 2015, the Islamic State made approximately $1 billion dollars in total revenue, which came equally from tax revenues and oil sales. Estimates for the first quarter of 2016 based on press reporting and events in Iraq and Syria suggest a significant drop in revenue to something under $700 million a year, with a further decline to about $440 million a year in July. While Treasury officials have routinely noted how difficult it is to come up with reliable figures, they have no incentive to systematically underestimate the group’s revenue; political risk would arise from making estimates that are too small. These figures can therefore be seen as a reasonable starting point. Other estimates of tax revenues are as high as $800 million in 2015.[c]
While they may seem large, these figures imply either an economy that has completely collapsed, an incompetent tax collection system, massive population outflows, or some combination of the three. This can be seen by comparing various revenue numbers to estimates of population and of pre-war economic activity.[d]
Screen capture from video released by al-Hayat in August 2015 entitled “Rise of the Caliphate and the Return of the Gold Dinar”
Combining the various publicly available maps estimating the regions the Islamic State controlled in 2015 with the two credible, scientifically validated sources of highly localized population estimates suggests that areas where the Islamic State can tax had a pre-war population of 2.8 million-5.3 million people.[e] The larger number assumes the group can tax in all areas where it has a substantial presence, while the smaller number assumes it can only do so in core territories.
Since the territory the group controls is not part of states that reliable reported sub-national productivity numbers, one can estimate pre-war GDP of the maximal definition of taxable Islamic State territory in two ways. First, one can use G-Econ, a widely-used geospatial database of economic activity that estimated economic activity in 1990 for each one degree by one degree grid cell of the globe, roughly 110-kilometer-by-90-kilometer cells in the Middle East.[f] If we assume productivity increases were constant across countries, which is clearly not precise and likely overestimates this region’s productivity as neither Iraq nor Syria notably outperformed world averages from 1990-2012, then we can assess the pre-war GDP of the Islamic State’s area by comparing the area to others that were close to its level in 1990.[g] The closest two countries in rank in G-Econ are Cameroon and Coté d’Ivoire, which had 2013 GDPs of $30 billion and $31 billion, respectively. Another way to assess pre-war economic activity in Islamic State territory is to use total nighttime illumination of the area it controls as measured from space as a proxy for economic activity.[h] Luminosity in 2012 for territory subsequently seized by the Islamic State was comparable to that of Ghana and Uruguay in that year, which had 2013 GDPs of $48.1 billion and $55.7 billion, respectively. Thus, we have pre-war GDP estimates ranging from $30 billion to $56 billion for a territory holding 5.3 million people. This corresponds to pre-war GDP per capita of the area between $2,021 and $3,774 depending on whether we choose the G-Econ or nightlight-based estimate.
Screen capture from video released by Wilayat Kirkuk in August 2016 entitled “And He Causes Charitable Deeds to Prosper”
Unfortunately, one cannot use direct measurement to assess how the economy has likely changed during the war. Luminosity clearly dropped as the war in Syria progressed and the lights basically went out in Iraq after the Islamic State took over. But the observed drop is a function both of economic woes and of the failure of the electricity grid, and we cannot take it as a clean indicator of economic collapse.
Instead, the best way to think about how the economy is doing in Islamic State areas is to compare pre-war GDP to current revenue estimates and ask what that would indicate about the group and the economy in areas it controls. As already noted, the U.S. Treasury estimated $500 million in tax revenue for the group in 2015. Start by assuming the group is average at taxation. The world median tax-to-GDP ratio is 17 percent. Thus, if the Islamic State is as effective at taxing the economy as a “normal” country and if no one moved out, then $500 million in taxation income would imply a per-capita GDP of $554-$1,050 in 2015 depending whether we assume the group is taxing a large or small area. That number represents a massive economic collapse.
Alternatively, if we assume the economy of the areas it controls has only shrunk by half, then the Islamic State is only achieving a tax-to-GDP ratio of 6.3 percent, meaning that it has an incompetent taxation system. Or if one assumes the group can tax at the world median and that GDP per-capita remains in the $3,000/year range (a figure between the two pre-war estimates), then $500 million in 2015 revenue implies a taxable population of only 980,400 people.
Put more starkly and as stated earlier, either the Islamic State is failing to tax, suffering economic collapse, facing massive population flight, or some combination of the three.[i] Either way, placing early revenue estimates into comparative perspective would have clearly indicated that the Islamic State was not on a path to building up an administratively competent state. This should not have been surprising; the group was starting from scratch, taking over already economically ravaged populations in an active war zone, and facing regular attacks on its personnel. Given the group’s goals of territorial conquest, however, those predictable failures should have clearly indicated how limited the Islamic State’s prospects were from the start.
Predicting the Islamic State’s Collapse
In late 2014 there were a few ways to think about the Islamic State. One way was as a terrorist group that had survived a long period underground and then achieved territorial control beyond anything prior jihadist groups have managed. Another was as a fledgling state trying to fight a three-front war (Iraq to the east, the Kurds to the north, and Syria and other insurgents to the west). Thinking about the group as the latter would have led to stark conclusions about its prospects given two simple meta-theories—one about what it takes to fight a conventional war for territory and another about the institutions required for a productive economy.
The meta-theory about what it takes to fight a conventional war for territory is a simple one. Creating combat power requires revenue because ammunition, equipment, and fuel cost money. While terrorism can be carried out on the cheap, holding territory is costly. And the Islamic State did not have the option of guerilla warfare; the territory it occupied was unsuitable to that approach. And if sustaining combat power requires a reliable source of revenue, then any group would be expected to fail unless it had stable outside funding, a viable tax base, or the ability to generate sustained revenues from natural resources.
While there is clear evidence that external support can enable rebel groups to stand against state forces for long periods, there is little evidence that the Islamic State has any major outside sponsors and the group’s own doctrine argues against reliance on external donors. And when it comes to the tax base, a second meta-theory comes into play that would predict failure for the group. Over the last two decades, the field of political economy has come to a broad consensus about the political and economic institutions required for fostering economic growth.[j] In particular, strong property rights, predictable taxation, functioning credit markets, and a clear regulatory framework are all necessary for economies to thrive. Unpredictable autocratic regimes typically suffer terrible economies, and states with high and unpredictable tax rates typically see their economies crumble over time.[k]
By early 2015 there was extensive reporting on the low quality of the Islamic State’s governing institutions and on the capricious nature of its regulatory and tax institutions. And by late 2015 reporting was widespread that the economy was in terrible shape. Given the theory about what was required in the abstract for an economy to be successful, it was foreseeable that the economy under the Islamic State’s control would fail over time, as it has.[l] Some posited that oil extraction could solve this shortfall. Others noted that the group’s oil infrastructure was inherently vulnerable to attack from the air, that it had to sell at a steep discount to world market prices, and that it likely lacked the ability to maintain fields for the long run. These arguments implied that over the long run the group could not make up its tax shortfalls with oil production. And while there are examples of dictatorial regimes around the world that have survived for decades, none have been fighting three-front wars.[m]
Even assuming away economic collapse, simple data-driven comparisons to other states in late 2014 boded poorly for the group. The Islamic State’s territory guaranteed that it would be a poor state compared to the ones it was fighting; in 2012 the nighttime illumination of territories it controlled at its peak amounted to no more than one-third of that in the rest of Syria and one-eighth of the rest of Iraq.[n] So what do we see if we assume the Islamic State would transform the economic activity it did have into military spending at rates similar to comparably sized states? Worldwide defense expenditures in 2014 peaked at 10.2 percent of GDP in South Sudan, with many conflict-affected countries only managing to spend 3 percent of GDP on defense. Assuming the Islamic State’s territory would maintain its pre-war GDP of approximately $30 billion, which was unlikely for the reasons given above, then defense expenditures observed elsewhere suggested the group could support military spending in the $900 million-$3 billion/year range. While we know now that it achieved nothing like that, even those numbers were tiny compared to Iraq’s 2014 spending of $9.5 billion, Turkey’s $20 billion, UAE’s $22.6 billion, or Saudi Arabia’s $80 billion, let alone the $8 billion in U.S. spending on the Islamic State campaign to date. While military spending does not translate directly into military power, the gap between what is financially feasible for the Islamic State over the long-run and what its neighbors spend would have led one to predict failure.
What Comes Next?
It would be tempting to think that the predictable economic dysfunction of the Islamic State would lead to collapse from within. That need not happen. The populations in Sunni areas under its control are extremely resilient. For more than a decade these populations have survived in a war economy. They have presumably developed a high toleration for economic pain as well as robust coping mechanisms. The history of Iraqi government mistreatment in 2013, Islamic State success in playing the sectarian card, and the group’s brutal repression of dissent means that economic misfortune is not necessarily going to translate (at least in the short- to medium-term) into political collapse.[o] But it will mean a steady degradation of combat power, one that cannot be compensated for with the capture of enemy equipment over the long run. Gear captured on the battlefield does not come with spare parts, much less the expertise required to maintain it, meaning that its value will steadily degrade with time, as it already has.
But the observation that the Islamic State will continue to steadily decline highlights an inherent limitation of theory in thinking about threats like the Islamic State. While political economy and some simple comparisons would have told us that the group was doomed to eventually fail—the economy it had to draw on was too small and its institutions inadequate given the task it set by fighting such a broad war—they are of little help in thinking about specific timing. The fact that the group is on the glide path to decline and collapse does not tell us when it will fail. Theory and comparison could have predicted the Islamic State was unlikely to be sustainable, but it did not tell us how long it would take to fail or what the final triggering event would be for the group’s territorial collapse.
Simple comparisons plus a bit of basic political economy could have predicted that the Islamic State would fail once its initial drive into Iraq stalled at the borders of the country’s Sunni-majority region. The broad lesson from this example is that estimates of groups’ potential should not eschew broad macro-theory. Rather, such a theory can help keep the terrorist threat in perspective.
And this last point is critical because as the Islamic State is driven back, history suggests we should expect increased terrorist attacks outside of the Middle East as the core of dedicated activists looks for ways to continue the fight despite losing territory. Many rebel groups shift into terrorism as their territorial ambitions are stymied. Indeed, in an earlier incarnation, the Islamic State did exactly that. And after the end of the Afghan jihad, foreign fighters looking for new struggles created new groups targeting other countries, including the United States. But just because a terrorist threat can carry out periodic attacks does not mean it can take and hold territory or foment revolution. Few terrorist groups succeed in doing so, and there is little theoretical reason to expect the Islamic State will be different.
Jacob N. Shapiro is Professor of Politics and International Affairs at Princeton University and co-directs the Empirical Studies of Conflict Project. His research projects examine political violence, economic and political development in conflict zones, and security policy. He is the author of The Terrorist’s Dilemma: Managing Violent Covert Organizations and co-author of Foundations of the Islamic State. He has served in the U.S. Navy and U.S. Navy Reserve.
This article draws heavily on previous work with Jamie Hansen-Lewis and Quy Toan-Do. Ben Crisman and Manu Singh provided excellent research assistance. The author wishes to thank Paul Cruickshank and the CTC editorial team for their considerable editorial input.
[a] The author first made many of this article’s arguments in print in Jamie Hansen-Lewis and Jacob N. Shapiro, “Understanding the Daesh Economy,” Perspectives on Terrorism, which was published in the fall of 2015. The first outline of that piece is dated May 25, 2015.
[b] The group was also referred to as the Islamic State of Iraq and Syria or the Islamic State of Iraq and the Levant (ISIL).
[c] This higher number is 33 percent of the $2.435 billion reported as “extortion” income in Jean-Charles Brisard and Damien Martinez, “ISIS Financing 2015,” Center for the Analysis of Terrorism, 2016. If one accepts such estimates, then rationalizing the reported numbers does not require one to think the economic collapse is quite as bad, the taxation quite so poor, or the population loss so great.
[d] The ideal date to use for pre-war economic activity is not obvious. In Iraq the shift to full-scale conflict did not happen until 2014. In Syria 2011 was the last year of relative normality in areas currently under Islamic State control. The author uses 2013 GDP figures here, but the results are not particularly sensitive to the year.
[e] These two sources, WorldPop and LandScan, combine census data with satellite imagery to estimate population numbers for every 800-mile-by-800-mile parcel of land on Earth. Both provide similar estimates for Islamic State territory.
[f] G-Econ provides a single snapshot in time of economic activity based on administrative data for 1990 and various other sources. It has not been recalculated for other years and so cannot be used to contrast the current economic situation with the state of the economy in the “caliphate” area before the international coalition launched operations against the Islamic State. On G-Econ, see Nordhaus, William, Qazi Adam, David Corderi, Kyle Hood, Nadejda Makarova Victor, Mukhtar Mohammed, Alexandra Miltner, and Jyldyz Weiss, “The G-Econ Database on Gridded Output: Methods and Data,” Yale University, 2006.
[g] Such an overestimate would inflate the level of collapse, but the fact the luminosity measure produces a higher GDP figure mitigates this concern.
[h] This analysis uses 2012 data from the National Oceanic and Atmospheric Administration’s (NOAA) National Center for Environmental Information (NECI) Version 4 Defense Meteorological Satellite Program’s Operational Linescan System Nighttime Lights Time Series (DMSP-OLS). These are the standard sources for stable nighttime illumination from satellites.
[i] Critically, though, using larger estimates would not substantially change the conclusions here. Even at the $800M in 2015 tax revenue the Islamic State would be taxing its territory at well less than 10 percent of any reasonable estimate of pre-war GDP. Again, it is either incompetent at raising revenue or its economy has collapsed or its population has fled or a combination of all three. None bodes well for long-run success.
[j] For a historically informed summary of the consensus, see Daron Acemoglu and James A. Robinson, Why Nations Fail: The Origins of Power, Prosperity, and Poverty (New York: Crown Business, 2012), especially chapter 3. The key point is that growth under extractive institutions is possible for a short period but not over the long-run.
[k] Zimbabwe is an excellent example of a potentially rich state destroyed by poor governance.
[l] It is difficult to evaluate how the Islamic State’s governance would have performed absent significant external pressure (e.g. airstrikes destroying both infrastructure, cash, and people as well as pressure on Turkey to tighten its borders against smuggling), which would have made it hard for even a legitimate and capable government to function. The broad empirical record suggests the Islamic State would have done poorly even absent those pressures. They likely sped up a process of immiseration that would have taken place in any case.
[m] Poor economic governance has proven survivable, but note that none of the regimes that survived it were routinely violating the Westphalian norm of sovereignty by attacking their neighbors and exporting terrorism around the world. Such actions rules out the kind of neglect by the international community that allows extremely low-quality rulers to survive over long periods.
[n] These numbers are based on aggregating illumination from grid squares the Islamic State currently controls in the DMSP-OLS data and comparing that to aggregated illumination in other areas of Iraq.
[o] Moreover, economic conditions are poor in areas of Iraq and Syria not under Islamic State control, and so residents comparing their economic welfare to that of neighbors in government-controlled areas may not be as inclined to rebel as they would be if those areas were doing better.
 On the latter, see Marc Sageman, Understanding Terror Networks (Philadelphia: University of Pennsylvania Press, 2004).
 On common organizational constraints, see Jacob N. Shapiro, The Terrorist’s Dilemma: Managing Violent Covert Organizations (Princeton, NJ: Princeton University Press, 2013). On trends in the demise of terrorist groups, see Seth G. Jones and Martin C. Libicki, How Terrorist Groups End: Lessons for Countering al Qa’ida (Santa Monica, CA: Rand Corporation, 2008).
 See, for example, Nour Malas and Maria Abi-Habib, “Islamic State Economy Runs on Extortion, Oil Piracy in Syria,” Wall Street Journal, August 28, 2014, and “Where Islamic State gets its money,” Economist, January 4, 2015.
 For a detailed version of this history, see Patrick Johnston, Jacob N. Shapiro, Howard J. Shatz, Benjamin Bahney, Danielle F. Jung, Patrick K. Ryan, and Jonathan Wallace, Foundations of the Islamic State: Management, Money, and Terror in Iraq (Santa Monica, CA: RAND Corporation, 2016), chapter 2. The author draws on that account here.
 See Craig Whiteside’s series of articles on the ISI campaign from 2009 to 2013 on War on the Rocks, particularly “War Interrupted, Part I: The Roots of the Jihadist Resurgence in Iraq” and “War Interrupted, Part II: From Prisoners to Rulers.”
 Congressional Testimony, Assistant Secretary of the Treasury Glaser, June 9, 2016.
 “Islamic State Monthly Revenue Drops to $56 million, IHS Says,” IHS Conflict Monitor, April 18, 2016.
 “Islamic State Caliphate Shrinks a Further 12 Percent in 2016, IHS Says,” IHS Conflict Monitor, July 10, 2016.
 Paul Cruickshank and Nicole Magney, “A View from the CT Foxhole: Adam Szubin, Acting Under Secretary for Terrorism and Financial Intelligence, U.S. Dept. of Treasury,” CTC Sentinel 9:8 (2016).
 For more details on the data behind these estimates and the assumptions that went into them, see Jamie Hansen-Lewis and Jacob N. Shapiro, “Understanding the Daesh Economy,” Perspectives on Terrorism, 2015.
 X. Li and D. R. Li. “Can Night-time Light Images Play a Role in Evaluating the Syrian Crisis?” International Journal of Remote Sensing 35:18 (2014): pp. 6,648–6,661; X. Li, R. Zhang, C. Q. Huang, and D. R. Li., “Detecting 2014 Northern Iraq Insurgency Using Night-time Light Imagery,” International Journal of Remote Sensing 36:13 (2015): pp. 3,446–3,458.
 Stathis N. Kalyvas and Laia Balcells, “International System and Technologies of Rebellion: How the End of the Cold War Shaped Internal Conflict,” American Political Science Review 104:3 (2010): pp. 415-429.
 Foundations of the Islamic State, p. 205.
 Timothy Besley and Maitreesh Ghatak, “Property Rights and Economic Development,” in Handbooks in Economics, Dani Rodrik and Mark Rosenzweig eds., vol. 5 of Handbook of Development Economics (Philadelphia: Elsevier, 2010), pp. 4,525–4,595.
 See, for example, Hanan G. Jacoby, Guo Li, and Scott Rozelle, “Hazards of Expropriation: Tenure Insecurity and Investment in Rural China,” American Economic Review 92:5 (2002): pp. 1,420–1,447.
 Dean Karlan and Jonathan Morduch, “Access to Finance,” in Handbooks in Economics, Dani Rodrik and Mark Rosenzweig eds., vol. 5 of Handbook of Development Economics (Philadelphia: Elsevier, 2010), pp. 4,703–4,784.
 On governance broadly, see Erika Solomon, “The ISIS economy: Meet the new boss,” Financial Times, January 5, 2015. On the unpredictability of tax institutions, see Ruth Sherlock, “Why business is booming under the Islamic State one year on,” Telegraph, June 8, 2015.
 See, for example, Joanna Paraszczuk, “The ISIS Economy: Crushing Taxes and High Unemployment,” Atlantic, September 2, 2015.
 On engineers, see Amy Myers Jaffe, “Commentary: ISIS victories and oil,” fuelfix.com, May 25, 2015. On bombing, see Maggie Ybarra, “Pentagon: Most of Islami State’s oil refineries in Syria have been destroyed,” Washington Times, September 30, 2014.
 Eli Berman and Jacob N. Shapiro, “Why ISIL Will Fail on Its Own,” Politico, November 29, 2015.
 This argument was first made in Hansen-Lewis and Shapiro, “Understanding the Daesh Economy.”
 Military Expenditures Database, Stockholm International Peace Research Institute (SIPRI), 2015.