The Economic Impact of Terrorism from 2000 to 2018

1.1 Conceptualising a Costing Model for the Impact of Terrorism

Estimating a monetary value for losses due to terrorism can be conducted in the same way as estimating the economic cost of violence. A large number of studies measure the economic cost of mortality and morbidity due to violence using a diverse range of methodologies. Iqbal, Bardwell, and Hammond (2019) suggest that the monetary cost of violence varies and is dependent on the underlying theory and assumptions used to estimate the costs. It is therefore necessary to provide an overview of the relevant approaches to costing violence and in particular, terrorism.

In generalising the literature that examines the monetisation of violence, four overarching theories can be identified, the bottom-up costing method also known as the cost accounting, contingent valuation, hedonic pricing, and econometric modelling of the variables and impacts of terrorism.

Dubourg, Hamed, and Thorns (2003), Iqbal, Bardwell, and Hammond (2019), McCollister, French, and Fang (2010) and Heek et al. (2018) employ the cost accounting approach to determine the costs of violence. This approach aggregates the direct and indirect costs of incidents of violence. McCollister, French, and Fang (2010) use this approach to estimate both unit costs and the total cost of crime for the United States. The paper aggregates tangible and intangible costs for victims, perpetrators and the public system such as the medical and criminal justice system costs. Iqbal, Bardwell, and Hammond (2019) use the unit costs provided by McCollister, French, and Fang (2010) to estimate the global economic impact of violence for 163 countries from 2008 to 2018. This methodology is applicable to determining the economic cost of terrorism.

Carnis (2004) states that the absence of a voluntary agreement between the victim and perpetrator results in the inability to formally price violent incidents given violence is not transacted in the market. The absence of a market mechanism to set the price or compensation for victims of violence creates a difficulty in estimating the cost of violence. Methods of contingent valuation are employed to estimate a monetary price for economic goods (economic bads in the case of violence). For example, the willingness to pay (WTP) is a technique of the contingent valuation method which enables the costing of different types of violence (Cohen et al. 2004). The benefits of using WTP method to measure violence is that it can provide an estimate of violence without the incident ever having taken place. The major difference between the contingent valuation and bottom-up approach is the former prices violence before the violence takes place, whereas the latter approach costs it after the incident. In other words, the bottom-up method calculates the cost on the question of ‘how much will it take to fix the impact’. Whereas contingent valuation costs based on the question of ‘what is the monetary value an individual or society is willing to pay to avoid the violence incident’ or alternately, ‘how much compensation is required if you were to suffer the violent incident’ (Iqbal, Bardwell, and Hammond 2019). This method could be used to estimate how much an individual or a society is willing to pay or invest in prevention programs to avoid terrorism (Mumpower et al. 2013).

Another approach to cost the economic impact of terrorism is to estimate the effect of terrorism on the price of goods such as housing in a particular area. This is achieved using the hedonic pricing methods that measure the price of specific goods by its characteristics. The best example is the price of real estate and how it is influenced by the number of rooms and other facilities, proximity to schools and hospitals and the level of crime in the area. Dubin and Goodman (1982) use the hedonic pricing approach to show that the presence of violence negatively impacts housing prices in Baltimore. Similarly, Tita, Petras, and Greenbaum (2006) use the hedonic pricing method in Columbus, Ohio, United States to find that violent crime has impacts on housing prices.

Another widely used approach is to measure the impact of terrorism on macroeconomic variables such as growth, investments and financial markets. Gaibulloev and Sandler (2008) find terrorism reduces the level of foreign direct investment as well as reduces gross domestic product growth rates (GDP). However, counter-terrorism strategies might increase aid flow and other interventions into a country. The US war on terror in Afghanistan is such an example where Afghanistan received increased support from the United States and other major international donors for the rehabilitation of the economy and government intuitions (Estrada et al. 2015). However, in the absence of peace, these achievements are extremely fragile. Yet, a contrast can be found in Zakaria, Jun, and Ahmed (2019) who show that a 1% increase in terrorism reduces foreign direct investment (FDI) by 0.1%, and economic growth by 0.002%.

Studies have used counterfactual scenarios to estimate direct and indirect costs by comparing states that suffer from violence to those that do not. The difference in the two outcomes is then assumed to be due to violence (Abadie and Gardeazabal 2003, 2008). Abadie and Gardeazabal (2003) show that in the case of Basque Country in Spain, GDP per capita declined by 10% during 1980 and 1990 relative to other states when the Basque region experienced terrorist activates. Bilgel and Kharasan (2017) estimates the difference in GDP per capita in Turkish states. States that suffered from terrorism experienced nearly a 14% decline in GDP over 21 years compared to states that did not suffer from terrorism. More recently, Sun and Yu (2020) found that Tibetan regions suffered a 27% reduction in GDP per capita due to separatist activities.

Pakistan suffers from a high level of terrorism activities that has led to a reduction in economic activity, created uncertainty and increased risk perception leading to decreased confidence (Ali 2010). Hyder, Akram, and Padda (2015) found a 1% increase in the number of terrorist incidents resulted in a reduction of per capita GDP growth by 0.39 percentage points. Estrada et al. (2015) estimate the total cost of terrorism in Pakistan to amount to $103 billion.

The adverse economic impact of terrorism varies based on the economic development of an economy, as well as the scale, impact and continuation of terrorist activities. Developed countries experience only short term impacts given the terrorism incidents are usually infrequent. Furthermore, developed economies are more resilient to short term disruptions. Gaibulloev and Sandler (2009) study the economic impact of terrorism in 42 Asian economies and find the impact is much smaller in the developed economies compared to the developing economies. When Gaibulloev and Sandler (2009) divide their model into developed and developing economies in Asia, they find that terrorism does not have a significant impact on the growth of the developed economies in the sample. However, among the 35 developing countries examined, for each additional transnational terrorist incident per one million inhabitants, the GDP per capita growth rate fell by 1.4%, and government spending as a percentage of GDP increased by 1.6%. Similarly, Gaibulloev and Sandler (2008) studied 18 Western European countries and find that for an additional terrorist attack per one million people, GDP per capita falls by 0.2% and the share of GDP directed to investment by 0.33 percentage points.

In addition to economic growth, terrorism also disrupts financial markets and trade, inhibits business investment and reduces tourism. Blomberg and Hess (2006), De Sousa, Mirza, and Verdier (2009), and Mirza and Verdier (2014) find a negative relationship between trade and terrorism. Blomberg and Hess (2006) estimate that terrorism, in addition to internal and external conflicts can distort trade as much as a 30% tariff on trade. Further, bilateral trade may be reduced by approximately 4% if one of the trading partner’s experiences domestic terrorism. De Sousa, Mirza, and Verdier (2009) find a similar reduction in bilateral trade between the United States and the countries where higher levels of terrorism occur. Nitsch and Schumacher (2004) and Bandyopadhyay, Sandler, and Younas (2017) confirm a reduction in bilateral trade and an increased cost of trade as a result of terrorism.

Terrorism increases the risk perception among investors leading to a decline in foreign direct investment (FDI). Abadie and Gardeazabal (2008) examine terrorism risks across 186 countries for 2003 and 2004 finding that the net FDI a country receives reduces by 5% of GDP in response to an increase in the terrorist risk by one standard deviation. Blomberg and Mody (2005) find transnational terrorism to have a greater negative impact on FDI to developing countries.

Tourism is particularly vulnerable to terrorism as it leads to heightened anxiety of travel. Enders and Sandler (1991) estimate the relationship between terrorism and tourism for Spain from 1970 to 1988 and find that on average, a typical terrorism incident results in 140,000 fewer tourists. Enders and Sandler (1996) quantify the dollar value reduction in tourism due to terrorism in a selection of European countries from 1974 to 1988, concluding that collectively, the countries lost $16.15 billion due to terrorism.

When estimating the global economic impact of terrorism, measuring both the direct and indirect costs that terrorism imposes on individuals and society is essential. Mayhew and Adkins (2003) identifies that the victim’s pain, suffering and lost quality of life should be included in the monetary cost of violence. Across the literature, this is defined as the intangible, or indirect cost of violence. Rajkumar and French (1997), Anderson (1999), McCollister, French, and Fang (2010) and Wickramasekera et al. (2015) are examples of papers that discuss or account for the direct and indirect costs of violence.

Direct costs of violence include the costs to the victim and perpetrator that can be monetarily measured. Examples of direct costs include hospital fees and forgone salary. However, the direct costs can include the widespread impacts, such as the costs to society such as policing, judicial systems and incarceration (Wickramasekera et al. 2015). The direct economic costs are the immediate costs to the victim, the perpetrator, and society that can be calculated monetarily in the market (Iqbal, Bardwell, and Hammond 2019).

Indirect costs extend beyond the immediate impact and do not involve a direct monetary exchange such as fear, pain, suffering, and lost quality of life (Wickramasekera et al. 2015). Indirect costs are often psychological and are therefore difficult to measure quantitatively. Estimating indirect costs requires the impact on society and the long term effects on individuals. Indirect costs are often undervalued using simple direct accounting methods as the costs do not have a specific market price; contrary to that of the direct costs.

An exhaustive systematic review of the literature on the economic cost of terrorism can be found in Gaibulloev and Sandler (2019) and Robinson et al. (2010). Wickramasekera et al. (2015) similarly provide a comprehensive literature review on the cost of violence, study designs and the crimes priced.

2.1 The Model

The economic impact of terrorism includes three components:

1. Direct costs – These are the cost of violence to the victim, the perpetrator, and the government. These include direct expenditure, such as the cost of hospitalisation.

2. Indirect costs – The costs accrue after the violent event and include indirect economic losses, physical and physiological trauma to the victim.

3. The multiplier – The multiplier represents the flow-on effects of direct costs, such as additional economic benefits that would come from alternate investment rather than from containing or dealing with violence.

Equation (1) below displays the formula for calculating the total economic impact of terrorism for country i in year t.

The multiplier for this study is equivalent to one. This indicates that every dollar saved from an absence of terrorism will be an extra dollar worth of economic activity from the flow-on effect. The multiplier only applies to the direct costs of terrorism and reflects the additional economic boost from reducing terrorism. The direct costs imposed by terrorism create additional productivity losses and to account for this loss, expenditure multipliers are commonly used in economic calculations (Brauer et al. 2009; Home Office 2011; Moretti 2010). The multiplier used in this paper is adopted from the multiplier published in Brauer et al. (2009). These authors use a multiplier equal to one to estimate the peace gross world product.

2.2 Data

Terrorism data is sourced from the Global Terrorism Database (GTD) which is collected and collated by the National Consortium for the Study of Terrorism and Responses to Terrorism (START) a Department of Homeland Security Centre of Excellence led by the University of Maryland. The GTD is considered to be the most comprehensive global dataset on terrorist activity and has now codified over 170,000 terrorist incidents. This dataset provides information on the number of deaths and injuries for each terrorism incident in addition to the level of property damage.

The property destruction estimates from the GTD are used to generate costs of property destroyed by various types of terrorist attacks. Each of the different property costs is further calibrated by country income type to adjust for the difference is property costs across different levels of economic development.

In order to capture the indirect effects of terrorism, GDP losses are included for countries that experience substantial impacts from terrorism; this is defined as over 1000 deaths within a year Collier (1999). There are a number of different ways to estimate the GDP losses from terrorism. Collier (1999), Polachek and Sevastianova (2012), De Groot, Bozzoli, and Brück (2015) and Costalli, Moretti, and Pischedda (2017) provide different methodologies and estimates of the impact of collective violence on the economy using different identification strategies and assumptions. The estimates of GDP losses from these studies range from 1 to 3% of GDP (Bozzoli, Brück, and Sottsas 2010). Collier (1999) use a large number of countries and concludes that 2.2% of GDP growth is lost for each year of war. Costalli, Moretti, and Pischedda (2017) estimate the GDP losses at 1.5%, which is the difference between GDP growth for countries with and without war. Bozzoli, Brück, and Sottsas (2010) provide a systemic review of the literature regarding the economic losses associated with conflict and large scale terrorism.

Abadie and Gardeazabal (2003, 2008) show that terrorism reduces GDP per capita by 10% compared to a synthetic control region. Bilgel and Kharasan (2017) estimates a decline of 13.8% in GDP per capita over a period of 21 years comparing states with and without conflict in Turkey. Ali (2010) estimates the cost of terrorism in Pakistan by suggesting that terrorism reduces short term economic activity leading to uncertainty and a reduction in confidence due to an increased risk perception. Enders and Olson (2012) suggest that the indirect cost of terrorism, including GDP losses, are larger than its direct costs. Gaibulloev and Sandler (2019) present an exhaustive review of the literature on the economic effects of terrorism.

At a 1000 deaths from terrorism, the scale of the violence will reach that of a civil war, and as stated by Dunne (2017), the differences among civil war, transnational organised crime, and terror groups are becoming less clear. Therefore, an assumption of this paper is when the scale of terrorism related deaths crosses the threshold of 1000 deaths in a year, GDP losses are to be included in the economic impact for that year. At this level, terrorism will create disruptions at the level of a civil war or an armed conflict. These losses are equal to 2% of the country’s GDP.

2.3 Converting Costs to Constant $US 2018

To compare the cost of terrorism between years, all costs are converted from current prices to constant using the GDP deflator. The GDP deflator data is sourced from the World Bank’s World Development Indicators and the International Monetary Fund World Economic Outlook. Where data is unavailable for all years for a particular country, but at least one year is available, we nominate to impute the missing data using the average between years.

Using 2018 as the base year for each country and dividing by each year for every country, we get the factor to multiply the unit costs by. When the unit costs are multiplied by the deflator factor, each country, for a particular year will have the unit costs in constant $US 2018. Equation (2) below displays the formula used for country i in year t.

The deflator factor is equal to one for each country in 2018. In order to control for outliers in the deflator, if a country’s deflator factor exceeds two, then we adjust their deflator to be equal to two for that particular year.

2.4 Scaling Unit Costs

Unit costs are used to estimate the cost of terrorism deaths and injuries. The unit costs from McCollister, French, and Fang (2010) are adjusted using the GDP per capita ratio as a scale. This is done using the ratio of GDP per capita in PPP terms. A country’s GDP per capita in PPP terms for a particular year is divided by the United States’ GDP per capita in PPP terms for the same year. Equation (3) below displays the formula used for country i in year t.

A country with a GDP per capita PPP that is 60% of GDP per capita PPP of the United States would have a terrorism death unit cost equal to 60% of the United States homicide unit cost.

In the absence of country-specific unit costs for all 163 countries, this paper uses adjusted unit costs. This adjustment attempts to correct for the differences in income levels taking the prices and purchasing power of each country into account. GDP per capita PPP is appropriate for the comparison of non-traded goods and services.

In an ideal scenario, this paper would estimate country-specific unit costs for different categories of crime and violence. These country-specific unit costs would then be used to calculate the aggregate costs of terrorism for each country. However, to develop the unit costs in a way similar to McCollister, French, and Fang (2010) for each country would require detailed country-specific data on wages, productivity, labour markets, hospitalisation costs and court compensation. In the absence of this data, calculating country-specific unit costs for all countries is difficult. As a result, the GDP PPP adjustment method allows for the comparison and scaling of 163 countries using a consistent approach and is therefore used in this paper to estimate a global cost of terrorism.

3.1 The Most Affected Countries

The highest level of terrorism in the 18 years to 2018 was recorded in countries that experienced armed conflict. These countries are mainly situated in the Middle East and North Africa (MENA), sub-Saharan Africa and South Asia. Since 2010, over 124,000 of the total 167,000 global terrorism deaths occurred in Iraq, Afghanistan, Nigeria, Pakistan and Syria. In other words, 74% of the global terrorism related deaths occurred in these countries. Figure 4 illustrates the terrorism death burden suffered by the above-listed countries.

Figure 4:

Number of deaths globally and by the five most affected countries, 2010–2018.

Afghanistan suffered the highest economic burden as a percentage of GDP of all the countries analysed in 2018. At 22% of GDP, the economic impact of terrorism is a significant drain on the Afghan economy. Afghanistan is one of the countries that has experienced rising levels of violence related to terrorist activities, contrary to the declining global trend. The economic impact of terrorism as a percentage of GDP is less than 5% in all countries except Afghanistan in 2018.

In 2014, the economic impact of terrorism was equivalent to 27% of Iraq’s GDP. Iraq suffered 3373 incidents and over 10,000 deaths from terrorism in 2014. Iraq’s two most drastic terrorist events took place in this year, the Sinjar massacre in Sinjar, Nineveh, Iraq and the Badush prison siege in Badush city, Nineveh, Iraq. The Sinjar massacre resulted in the death of at least 953 people and over 5000 people were abducted. This attack was attributed to the Islamic State of Iraq and the Levant (ISIL). The Badush prison siege saw the assailants release the Sunni inmates and kill 670 Shia prisoners and was similarly attributed to ISIL.

Iraq had the second-highest economic cost of terrorism as a percentage of GDP, equivalent to 4% of GDP in 2018. The 10 countries most affected by terrorism since 2010 are shown in Table 3 and all were experiencing ongoing conflict at the time of the study. This suggests that armed conflict and fragility provides a platform for terrorist groups to organise and undertake violent activities. An extension of Table 3 which includes more countries is available in Appendix.

3.2 Regional Trends in Terrorism and Its Economic Impact

The Middle East and North Africa is the most affected region by terrorism, followed by South Asia. Since 2000, MENA has experienced over 35,000 terrorist attacks that have resulted in over 95,000 deaths. Figure 5 shows the number of terrorist attacks and deaths by region aggregating the number of deaths and incidents from 2000 to 2018. Central America and the Caribbean is the least affected region suffering 225 attacks since 2000. These attacks have resulted in 190 deaths.

Figure 5:

Attacks and fatalities from terrorism by region, 2000–2018.

Figure 6 displays the disproportionate burden of terrorism deaths suffered by sub-Saharan Africa, the Middle East and North Africa and South Asia. These three regions have suffered 91% of all terrorism deaths since 2000. In 2014, MENA suffered over 14,000 terrorism deaths, the largest annual number of deaths of any region. In 2018, MENA suffered 2407 terrorism deaths, an 83% decline from its 2014 level.

Figure 6:

Number of deaths globally by region, 2010–2018.

The economic impact of terrorism in sub-Saharan Africa amounted to $US 12.2 billion in 2018, which is 37% of the global impact of terrorism. This is followed by the Middle East and North Africa where the economic impact of terrorism was $US 11.9 billion. As shown in Table 4, Central America and the Caribbean were least affected by the economic impact of terrorism, representing $US 120 million. This amounts to 0.4% or of the global economic impact of terrorism. In 2018, South Asia suffered the third largest economic impact driven by Afghanistan’s high economic impact of terrorism.

The following analysis examines the economic impact from terrorism deaths, injuries and property damage by attack type and by region shown in Table 5.

Bombings and explosions account for 66% of MENA’s economic impact from terrorism. It indicates that guerrilla tactics are used by terrorist groups in the region. Similarly, bombings and explosions were the costliest tactics for South Asia at 60% of its total economic impact by attack type, followed by armed assault at 31%. Sub-Saharan Africa is most impacted by armed assault at 65% of the economic impact from deaths, injuries and property damage. This analysis has strong implications for counter-terrorism policies in specific regions.

Similar to the regional terrorism attack tactics, countries of particular income levels are also more or less targeted by specific types of terrorism tactics. Across all income groups, armed assault, and bombing/explosions account for more than 50% of the economic impact of deaths, injuries and property damage as displayed in Table 6. This highlights the needs to promote policies that reduce armed assault terrorism tactics such as disarmament and reduces bombings by means of early bomb detection.

The economic impact of terrorism deaths, injuries and property damage in high-income countries is mainly armed assault at 45% of the total. Whereas, in low-income countries bombing and explosions make up 55% of the economic impact of terrorism deaths, injuries and property damage.

3.3 The World’s Largest Terrorist Attacks in the Last 18 Years

Table 7 highlights the 20 costliest terrorist attacks since 2000. This analysis excludes property damages and GDP losses and rather reflects the cost of death and injuries from these incidents. An extension of Table 7 is provided in the Appendix with the 100 costliest terrorist attacks since 2000.

The attacks of September 11, 2001 in the United States are the most significant attacks both in terms of deaths and injuries, but also in terms of economic impact. The economic impact of the attacks is estimated at $US 40.6 billion. The events of 9/11 led to 2957 deaths and almost 16,500 injuries. Kunreuther, Michel-Kerjan, and Porter (2003) estimate the cost of the 9/11 attacks in the United States at $US 80 billion, which is double the amount calculated in this study. The differences arise from what elements of the cost are included in the estimation. This study includes the cost of deaths and injuries as well as property loss. Other studies have included the cost of the insurance providers, the cost of business interruptions and worker’s compensation (Kunreuther, Michel-Kerjan, and Porter 2003).

Iraq recorded 31 of the 100 costliest terrorist attacks in the last 18 years, the most of any country globally. Those 31 incidents in Iraq combine to cost $US 31 billion in lost GDP from deaths and injuries. Iraq’s most severe incident was the Sinjar massacre. This event was the second-costliest incident globally from 2000 to 2018 in terms of deaths and injuries, resulting in an economic impact of $US 4.3 billion. This attack was attributed to ISIL in which the assailants attacked Yazidi civilians in Sinjar, Nineveh, Iraq. This event led to the deaths of at 953 people, and 5350 people were abducted in the assault.

Nigeria suffered the second-largest number of incidents with 14 incidents in the 100 costliest. Iraq and Nigeria have been among the 10 most affected countries by terrorism, with Iraq becoming the most affected country 14 times between 2000 and 2018. Russia and Egypt are the two other countries that appear in the 20 deadliest terrorist attack list. Russia’s costliest terror incident was a siege of a school in the town of Beslan, Russia. This event was the fourth costliest event and is estimated at $US 2.1 billion and resulted in 344 fatalities and 727 injuries.