Qual Quant DOI 10.1007/s11135-015-0179-z

How terrorism affects the economic performance? The case of Pakistan Alam Khan • Mario Arturo Ruiz Estrada • Zarinah Yusof

 Springer Science+Business Media Dordrecht 2015

Abstract We examine application of the Economics of Crime Monitoring Model (ECMModel) Ruiz Estrada and Ndoma (J Policy Model 36:867–882, 2014) on Pakistan terrorism activities. The application of ECM-Model is used to evaluate the impact of terrorism on the economic performance of Pakistan economy. First part of the research work is related to introduction and background of study. The second section is associated with the theoretical and conceptual frame work that explains how terrorism affects the economy. The third part of the paper describes the methodology of the model. The fourth part of this research paper elaborates the results of the study. The second last part of the paper is the econometrics techniques and results which support the model and last part of the research work is conclusion and recommendations. Keywords

Economic performance  ECM-Model  Terrorism  Pakistan

JEL Classification D74  H56

1 Introduction Many prominent economists like Keynes (1919), Pigou (1940) and Robbins (1942) had started writing about war, peace and day to day economic situation and they used economic models to device policies. Terrorism has got more attention in the economic literature as

A. Khan (&)  M. A. R. Estrada  Z. Yusof Faculty of Economics and Administration (FEA), University of Malaya, 50603 Kuala Lumpur, Malaysia e-mail: [email protected] M. A. R. Estrada e-mail: [email protected] Z. Yusof e-mail: [email protected]

123

A. Khan et al.

compared to internal and external conflicts after 11 September 2001 (Blomberg et al. 2004a, b). There are different views in literature about the relationship between terrorism ¨ cal and Yildirim (2010), and economic performance. According to Gupta et al. (2004), O Sultan (2013) terrorism is the main obstacle in the way of economic growth in developing economies. Some scholars like Meierrieks and Gries (2012) examined that terrorism and economic growth has no link. Terrorism in Pakistan has more magnitude in terms of events and effects than most of the developed and developing economies like Israel, Greece, Turkey, Spain and Unites States of America (USA) and continuously in world breaking news (GTD 2013). However, terrorism in Pakistan has got comparatively less attention in academic circle than other economies of the world. Very few studies have examined the effects of terrorism especially with respect to terrorism-economic performance side. Pakistan has experienced notable fluctuations in economic growth and terrorist activities during the last decade. Many high profile terrorist groups are operating in this region and Osama Bin Laden (the Al-Qaida leader) had been arrested and killed in Pakistan, so a strong link between economic growth and terrorism is extremely intuitive. This research work studies the impact of terrorism on economic performance in Pakistan economy. The terrorism activities can take place any time anywhere so it needs great care to quantify its effects on economy of Pakistan. The ECM-Model is suitable technique for indicating how the gross national product (GNP) growth rate is directly related to the incidence of terrorism.

2 Background of study Among other factors, economic growth and international trade possess pivotal position in economic health of any country. Likewise, improvement in economic growth and trade raise the living standard of people subduing poverty. Furthermore, growth of a country depends heavily upon labour force, physical infrastructure and technological progress, both in term of quantity and quality. Similarly, demographic condition of the economy, investment in physical infrastructure and improvement in institutional quality influences international trade of a country significantly (Abel et al. 2008). Additionally, factors like natural hazards, natural disasters, civil wars, internal conflicts, external disputes with neighbours countries and terrorism can also influence the economic growth and international trade of any economy (Gupta et al. 2004). With all, terrorism reduces economic activities by destroying the human and physical capital which are considered as the main inputs of production. The decrease in inputs is related with the reduction in domestic output and as a result economic growth is badly affected. The economic growth is also affected indirectly by terrorism through the allocation of domestic savings, investment decisions and resources. These factors not just reduce the local economy production, but also affect negatively the exports of that economy and international trade. Buckelew (1984, p. 18) define terrorism as ‘‘violent, criminal behaviour designed primarily to generate fear in the community or in a substantial segment of the community for political purposes’’. FBI (2012, p. 1) defining terrorism as ‘‘Terrorism is the unlawful use of force or violence against people or property to intimidate or coerce a government, the civilian population, or any segment thereof, in furtherance of political or social objectives’’.

123

How terrorism affects the economic performance?

2.1 An overview of Pakistan Pakistan is located in South East Asia. The neighbour countries are Afghanistan in the West, China in the North, India in the East and Iran in the South West. Pakistan has the second largest border with Afghanistan after India. The total are of Pakistan is 796,095 Km2. The total population of the economy is 188.2 million, which is the sixth largest country of the world by population. There are different ethnic groups like Punjabis, Balochistanis, Pashtuns, Sindhis, Muhajirs, Saraikis, Hindkows and Chatralis. More than 97 % of the population is Muslim. Minority groups are Sikhs, Christians, Parsi and Hindus. The per capita income was US $1386 in year 2013. Gross domestic product (GDP) growth rate in 2013 was 4.14 %. The services sector is contributing 58.1 % of GDP, while agricultural sector is sharing 21 % of GDP to the total economy. The total labour force of Pakistan is 59.74 million in 2013. Agriculture absorbs 43.7 % of labour force and manufacturing sector labour force participation rate was 14.1 % in a year 2013. The total exports of the Pakistan economy in 2013 was 20.143 billion US dollar while imports during this time was 36.66 billion dollar. The trade deficit during 2013 was $16 billion (Economic Survey of Pakistan).1 2.2 Evolution of terrorism in Pakistan Pakistan is located in such a geographical position where on one side there is Kashmir dispute with India and on the other side the border with Afghanistan is not secured. Pakistan has faced the terrorism issue since 1947, but in 1980s the issue of terrorism was restarted with the Afghan war. Most of the militants from different regions of the world came to Pakistan to fight against the Russia. This was foundation and the root cause of the terrorism in Pakistan where these militants were set together in the North part of the Pakistan (Weiner 1998). After the Russian defeat in Afghanistan, the control of Afghanistan came in the hands of Taliban. The 9/11 incident of 2001 in US has brought new dimension in terrorism and attacked on Afghanistan, where the Pakistan has provided the logistic support to America against Afghanistan Taliban. Due to which the militancy and terrorist activities started in Pakistan. The economy of Pakistan has suffered too much due to terrorism, which cost the economy in the form of more than 35 thousand citizens, 35 hundred armed forces, domestic migration of millions of people from one part of country to another part of country for shelter, destruction of infrastructure, erosions of investment environment, reduction on production and increase in unemployment. Due to this chaotic situation, the economic growth has slowed down and inflows of foreign investment have automatically badly been affected, stressed by the travel bans issued by western countries to its entrepreneur. Pakistan economy has paid the price of terrorism in terms of both economic and security. The economy is subjected to huge direct and indirect cost which increased from $ 2.669 billion in 2001–2002 to $ 13.6 billion. The terrorism has badly affected the exports, stopped the inflows of foreign investment, shrunk import demand, decreased tax collection, increased in security spending and reduced the local as well as foreign tourism. Pakistan’s investment to GDP ratio has decreased from 22.5 % in 2006–2007 to 13.4 % in 2010–2011 (GoP 2013). According to the Global Terrorism Database (GTD),2 in the year 2013, the number of people died in terrorist attacks almost 2891, while as compared to the year 2000 it was 1

Various issues of economic survey of Pakistan.

2

http://www.start.umd.edu/gtd/.

123

A. Khan et al.

around 118. This drastic increase in the death polls occurred due to the increase in the number of suicide attacks. GTD source further explained that there were 86 suicide bomb attacks in Pakistan in 2013 as compared to only two attacks in 2002. From 2004 onward the terrorist attacks affect at large proportion. The evidences of post attacks have shown that almost young children of 14–17 were involved in those suicide bomb attacks. This means that all those young children were financially deprived. The summary of terrorism activities of Pakistan are shown in (Fig. 1). 2.3 The effects of terrorism on economic growth in the case of developed countries Meierrieks and Gries (2012) studied the relationship between terrorism and economic performance in 18 Latin American economies. The main variables of the study are total terrorism attacks, real GDP per capita and control variables are political instability and military spending. The time period of the study is 1970–2007. Panel Granger causality technique is used for the analysis of the study. The results of the study confirm that terrorism has no causal effect on the economic growth of the higher developed countries Latin America. This study concludes that economic growth and terrorism has no causal link in the higher developed countries of the said region. Gries et al. (2011) tested the causal linkages between domestic terrorism and economic growth in seven western European countries that are France, Germany, Greece, Italy, Portugal, Spain and United Kingdom (UK). The authors have used time series data from 1951 to 2004 for the examination of the study. Hsiao (1979, 1982) used granger causality technique to examine the casualty between terrorism and economic growth. The main variables of the study are real GDP per capita and domestic terrorism. Trade openness is the control variable of the study. The findings of the study show that economic growth leads terrorists violence in all countries of the study, while terrorism causality influence economic growth only for one economy that is Portugal. The authors have also mentioned future research study directions such as for this relationship includes other part of the world like Middle East, specific period of Time, specific kind of terrorism and a certain targeted aims like particular citizens. Total Terrorism events

12000 10000 8000 6000 4000 2000 0 2000

2001

2002

2003

2004

2005

2006

2007

2008

Total Terrorism events Fig. 1 Total terrorism events

123

2009

2010

2011

2012

2013

How terrorism affects the economic performance?

Blomberg et al. (2004a, b) studies the relationship between economic conditions and terrorism in 130 countries. The authors have used panel data from 1968 to 1991. The economic variables of the study are GDP per capita and investment. The static model of Grossman (1991) and the dynamic model of Tornell (1998) have applied for the analysis. The results of the study confirm that economic activities and terrorism are interdependent. They further said that high income countries have chances of more incidents of terrorist activities. The conclusion of the study is that terrorism has linked with business cycle, period of poor economic performance increase the possibility of terrorist activities. Blomberg et al. (2004a, b) said that in the organization of economic cooperation and development (OECD) developed economies, the terrorism have a very small and insignificant effect on economic growth. The authors of the study further said that internal conflicts have more drastic affect than external shocks in OECD countries. 2.4 The effects of terrorism on economic growth in the case of developing countries Shahbaz et al. (2013) examined with the title ‘‘Linkages between inflation, economic growth and terrorism in Pakistan’’. The time period of the study is from (1971) to (2010). He has used the autoregressive distributed lag (ARDL) bounds and rolling window approach for the analysis of the study. The variables of the study are inflation, economic growth and terrorist’s attacks. The results of the study explain that there is bidirectional causality between inflation and terrorism. The results also show that economic growth is also granger caused by terrorism and inflation. The author has given the recommendation that government should control inflation because it does not just affect the economic growth but also generates terrorism. Meierrieks and Gries (2012) studied the relationship between terrorism and economic performance in 18 Latin American economies. Among those 18 Latin American economies, the findings of the study show that economic growth has reduced terrorism in less developing countries of Latin America. But terrorism has no causal effect on economic growth of those developing countries (Guatemala, Ecuador) of Latin America. These findings of the study explain that economic growth can cause terrorism but terrorism cannot cause economic growth in the less developing countries of Latin America. ¨ cal and Yildirim (2010) studied the relationship between terrorism and economic O growth with a title ‘‘Regional effects of terrorism on economic growth in Turkey’’. The authors have used a geographically weighted regression (GWR) approach to estimate the parameters of the variables. The advantage of this technique over other techniques is that this model shows the relationship among variables that change over space by launching distance based weights to estimate the parameters for each and every variables and geographical location. The other models have ignored the socio-economic condition of the countries by doing cross country analysis. The time period of the study is from 1987 to 2007. The empirical results of the study have shown the negative association between economic growth and terrorism in Turkey. This study explains that terrorism negatively affects the economic growth of the economy of Turkey. The South-Eastern part of the Turkey is more affected by terrorism than Eastern-West area of Turkey because more terrorist activities have taken place in South-Eastern part of the country. Blomberg et al. (2004a, b) examined the macroeconomic consequences of terrorism in Africa, Asia and Middle East countries. The authors have selected GDP per capita, economic growth, trade openness and Gini coefficient as economic variables. They have found that statistically terrorism have a strong but economically low negative impact on economic growth. The effect of terrorism in non-democratic and African countries is too much

123

A. Khan et al.

large. The possible reasons which affect these economies are severe internal conflicts which have created external conflicts and terrorism. The results also explain that internal and external conflicts like wars have more effect on economic growth than terrorism. The authors have recommended a strong public policy to mitigate the all possible risk related to terrorism. 2.5 The effects of terrorism on international trade performance Walkenhorst and Dihel (2002) wrote that there is an increase in security measures due to increase in terrorism events. Because of these security measures and regulations, the international trade becomes difficult and more expensive. These obstacles of international trade increase the delivery time of traded goods. Terrorism increases the risk of destroying the traded goods. The terrorists target the means of transportation which is more vulnerable to the supply chain of specific transport modes (Nitsch and Schumacher 2004). Blomberg and Hess (2006) used the gravity model to study the effects of terrorism on bilateral trade. They used annual panel data from the year 1968 to 1999. Terrorism is measured through dummy values. The international trade is measured by the sum of imports and exports divided by two. The results confirm that bilateral trade is decreased by 5–6 % if at least one terror activity is occurred.

3 Theoretical and conceptual framework of study The main hypothesis of the study is that terrorist activities lead to poor economic performance. The justification of terrorists to destabilize the economy can be explained from the rational choice theory. A government that is attacked by terrorists may take on a rational perspective, where it compare the cost of accepting the terrorist demand against the cost of prolonged the terrorists campaign that results from continuous resistance by the government (Sandler and Enders 2008). Thus, destabilization is a main objective of terrorists. Poor economic performance due to terrorism means that accepting terrorists demand has become relatively less costly from the government perspective. Terrorism is anticipated to destabilise the economy by imposing direct cost and by doing actions that result in poor economic performance. The direct cost includes loss of human capital and the loss of property and infrastructure. The indirect cost comes from the response of economic actors to terrorism. Due to terrorism, security measures may be taken, which increases the transportation cost and obstacles in international trade.

4 An introduction to Economics of Crime Monitoring Model (ECM-Model) The ECM-Model (Ruiz Estrada and Ndoma 2014) explains that every economy is exposed to terrorism at any time and any place. Here in this application of the model crime is replaced with terrorism. The potential damage and impact of terrorism on GDP cannot be underestimated and ignored. To understand the dynamics of terrorism is not enough, but a proper evaluation of its quantification is significant for future policy recommendations. The ECM-Model is used for explaining that how GDP growth is directly related to the incidents of terrorism. The ECM-Model has five indicators (i) The total terrorism

123

How terrorism affects the economic performance?

frequency rate (b) (ii) The national terrorism vulnerability rate (lT) (iii) The terrorism devastation magnitude rate (k) (v) The economic desgrowth rate (d) (vi) The terrorism vulnerability surface (VV-surface). These five indicators of ECM-Model explain the various levels of vulnerability and devastation creating from various terrorism activities. The five indicators are calculated by assembling historical data of various terrorism activities. The terrorism events impact any economy in the form of destruction and loss of infrastructure capital and human capital. According to the ECM-Model, the economic impact of terrorism event includes decrease in production and human capital (labour force). In the ECM-Model, a new concept is introduced called economic desgrowth. The economic desgrowth explains the leakage of economic growth due to terrorism. The main purpose of the economic desgrowth is to determine the possible impact of terrorism on the final GDP growth rate over a period of time. The data used by the ECM-Model is based on the categories of the twelve terrorism activities. These activities include suicide, assassination, hijacking, kidnapping, barricade, bombing, unknown, armed assaults, unarmed assaults, infrastructure, number of killings and number of wounded. 4.1 The national terrorism vulnerability rate (lT) To calculate the total terrorism frequency rate (bi) is the ratio of a specific terrorism event in particular divided by the cumulate total frequency of the same terrorism event over the past decade including the current year (see Eq. 1). bi;t¼T bi ¼ Pt¼T t¼T9 bi;t¼T

ð1Þ

The value of bi lies between 0 and 1. 0  bi  1

ð2Þ

It supposes that terrorism can take place anytime in any part of the world. In this study we have twelve terrorism events frequency rates that are as follow, suicide (b1), assassination (b2), hijacking (b3), kidnapping (b4), barricade (b5), bombing (b6), unknown (b7), armed assaults (b8), unarmed assaults (b9), infrastructure (b10), number of killings (b11) and number of wounded (b12). Each terrorism activity has its own intensity according to its location and social dispute issues. In the ECM-Model it is supposed that each terrorism activity has its own nature and cannot be predicted with accuracy. To calculate the national terrorism vulnerability rate, the formula is as follow. The evaluation of the terrorism vulnerability rate (lT) is grouped into three various level of vulnerability (see Eq. 3). pffiffiffiffiffiffiffiffiffiffiffi ð3Þ lT ¼ ðLn 1  bÞ Level 1: High vulnerability ðRed colourÞ: 10:75 Level 2: Average vulnerability ðYellow colourÞ: 0:740:34 Level 3: Low vulnerability ðGreen colourÞ: 0:330

ð4Þ

123

A. Khan et al.

4.2 The terrorism devastation magnitude rate (k) To measure the crime devastation magnitude rate, two variables including capital devastation and human capital devastation are used. Capital devastation means total incidence of terrorism area in a geographical location divided by the total area of the same geographical location. Human capital devastation means the number of killed or missing persons divided by the total population of the same area of location. After getting the values of both variables, multiply both the results of capital and human capital devastation to find the value of terrorism devastation magnitude rate (k). k ¼ Ln½ð/kÞ  ðwLÞ

ð5Þ

4.3 The economic desgrowth Economic desgrowth (Ruiz et al. 2014) is defined as a macroeconomic indicator that explains the final impact of any type of any natural hazards on the GNP. This means that final GNP-post violence hazards effect depends on the terrorism devastation magnitude rate (k). At the same time, terrorism devastation magnitude rate (k) is directly proportional to the national terrorism vulnerability rate (lT). Thus, economic desgrowth is the product of lT and k. the formula of economic desgrowth are as follows: d ¼ ð lT ÞðkÞ

ð6Þ

The Eq. (6) explains that the value of economic desgrowth must be negative. The empirical analysis investigates that when lT and k increases, economic desgrowth will follow the same behaviour. " d ¼ ð " lT Þð" kÞC

ð7Þ

# d ¼ ð # lT Þð# kÞ

ð8Þ

So the above expressions conclude that economic desgrrowth is directly related to terrorism devastation magnitude rate and national vulnerability rate. 4.4 The terrorism vulnerability surface (VV-Surface) To draw the VV-Surface, it depends on the terrorism frequency rate (bi) values and megasurface coordinate space. The terrorism vulnerability surface is a three by four matrix that includes the single values of all twelve variables. The twelve variables values are drawn on three by four rows and columns on the VV-Surface. The VV-Surface is a bird eye view pictorial representation of the overall any hazards of any economy. The expression of VVSurface can be represented as below. g ¼ ðb1 ; b2 ; b3 ; b4 ; b5 ; b6 ; b7 ; b8 ; b9 ; b10 ; b11 ; b12 Þ

ð9Þ

The VV-Surface depends on the change that occurred due to any hazards during a particular period of time.

123

How terrorism affects the economic performance?

5 Results of the study 5.1 The application of the ECM-Model on the Pakistan terrorism events Using the ECM-Model in the case of Pakistan provides useful information about terrorism effects on economic growth. Before going into the details of the application of ECM-Model for Pakistan, there is a need to explain further the terrorism situation in Pakistan. The major terrorism events occurred in Pakistan are suicide attacks, bomb blasting, kidnapping, hijacking, killing and assassinations. Most of the terrorist incidents took place in Khyber Pakhtunkhwa, North Part of Pakistan and tribal areas. This area is more affected due to terrorism because of its border with Afghanistan. The other parts of a country which are affected due to terrorism are Karachi, Sindh and Quetta, Baluchistan. 5.2 The total terrorism frequency rate (bi) In this part of the research work, we attempt to investigate the terrorism vulnerability propensity rate in Pakistan. Figure 2 explains the terrorism growth rates (bi). Mainly Pakistan has a large intensity of terrorism events on its historical record. We distribute all the terrorism activities into three different colours that affect country on the basis of their terrorism growth rates (bi). Figure 2 also explains that highest terrorism activity rates in Pakistan is by unarmed assaults (b9) at 0.92. The second highest terrorism activity rate is hijacking (b3) at 0.84. The third place of terrorism activity rate according to its intensity is number of wounded (b12) at 0.79. Figures 3 and 4 show the terrorism vulnerability surface in twelve terrorism activities such as suicide (b1), assassination (b2), hijacking (b3), kidnapping (b4), barricade (b5), bombing (b6), unknown (b7), armed assaults (b8), unarmed

Fig. 2 Conceptual framework

123

A. Khan et al.

Fig. 3 Analytical framework

Fig. 4 The total terrorism frequency rate

The total Terrorism frequency rate

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

assaults (b9), infrastructure (b10), number of killed (b11) and number of wounded (b12). All these results show that Pakistan has faced a major problem of terrorism attacks in the region (Figs. 5, 6).

123

How terrorism affects the economic performance?

Fig. 5 Vulnerability surface

Fig. 6 Terrorism vulnerability surface

5.3 The Pakistan terrorism vulnerability rate (XT) We investigated a big difference margin between the maximum and minimum terrorism vulnerability rate (XT) for Pakistan. According to the terrorism historical data, Pakistan vulnerability rate ranges from (XTmin) = 0.12 to (XTmax) = 0.94. 5.4 The terrorism devastation magnitude rate (p) Here in this section, we would compare the terrorism devastation magnitude rate (p) of Pakistan between 2000 and 2013 to explain the effects of devastation on the economy. According to our estimated results, the devastation resulting from the 2000 was very limited at -0.09. In 2013, the devastation was noted at much higher at -2.62. We could

123

A. Khan et al.

conclude from the results that terrorism devastation magnitude rate (p) in 2013 produced more devastation than the terrorism devastation rate (p) in 2000. 5.5 The results of economic desgrowth (d) Lastly, we want to measure the impact of terrorism on economic growth; we use the novel concept of ‘‘economic desgrowth (d)’’, introduced by Ruiz Estrada and Ndoma (2014). The use of economic desgrowth helps to investigate potential leakages that can badly affect GNP performance. During the GNP construction, many leakages may develop due to various activities of terrorism events. According to our results, the economic desgrowth caused by terrorism in Pakistan has an impact of -0.04 % in the year 2000. The economic desgrowth created by terrorism is bigger in 2013 at -1.71 %. It is concluded that the difference between these two time periods explains -1.67 % according to our results estimates. The relationship between economic desgrowth and terrorism frequency rate has also shown in Fig. 7.

6 An econometric analysis of terrorism in Pakistan We want to support our results with econometric analysis for the relationship between terrorism and economic growth in Pakistan. The data for terrorism and economic growth have been selected from the year 1974 to 2010. The terrorism score is calculated from Global terrorism database which include the number of incidents, deaths, injuries and property damage. The economic growth rate is used as a dependent variable for the economic performance of Pakistan. We use the time series methodology for the econometric analysis of the study. First, we investigate the unit root tests to explain the order of integration of said variables. We also apply the Granger causality to explain the causation among variables. To explain the long-run and short relationship co-integration technique will be used and VECM model will be applied to separate the long-run and short-run coefficients.

Fig. 7 Economic desgrowth and terrorism frequency rate

123

How terrorism affects the economic performance?

6.1 Unit root tests In this study we are using time series data and such data requires to be checked for stationarity. The unit root test identifies is used to test stationary. Yt ¼ qYt1 þ 2t

ð10Þ

Here Yt is time series variable at time t, Yt1 is lagged time series of y, T is time, where t = 1.2.3… q is the coefficient of the lag of time series 2t is disturbance term which is supposed to be identically independent scattered with zero mean and variance (r2). When the q = 1 or q [ 1, it means that the data is non-stationary and if the q is less than 1, it represents the data is stationary. Yt  Yt1 ¼ qYt1  Yt1 þ 2t

ð11Þ

Yt  Yt1 ¼ DYt  2t

ð12Þ

Or we can write the new equation as following: DYt ¼ dYt1 þ 2t ;

ð13Þ

where d ¼ ðq  1Þ and D, as usual in the 1st difference method. On the 1st difference the data will be stationary, a series Yt is integrated of the order one or contain unit root, if the Yt is non-stationary but DYt is stationary. Dickey and Fuller (1979, 1981) devised a technique to properly test for non-stationary. The main intuition of their testing is that for non-stationary is corresponding to testing for the existing of a unit root. Yt  Yt1 ¼ bYt1  Yt þ 2t

ð14Þ

DYt ¼ b1 þ dYt1 þ 2t

ð15Þ

DYt ¼ b þ dYt1 þ 2t

ð16Þ

DYt ¼ b þ b2 þ dYt1 þ 2t

ð17Þ

It is supposed that the error term (2t ) is not correlated. If the error term (2t ) is correlated, Dickey and Fuller have introduced a test identified as Augmented Dickey Fuller test (ADF). The augmented test is consisting of the subsequent regression equation. m X DYt ¼ b0 þ b1t  1 þ biDYt1 þ 2t ð18Þ i¼1

DYt ¼ b1 þ b2t þ dYt1 þ

m X

/ DYt1 þ 2t

ð19Þ

i¼1

The idea to include sufficient terms is that the error term in equation is serially uncorrelated. 6.2 Granger causality test This technique is introduced by Engle and Granger (1987). This explains that if the data which are not stationary at their levels. The data at stationary in the first difference can be modelled with their level conditions. It is probable to exam how several co-integration

123

A. Khan et al.

vectors are between the variables by applying the VAR. Though, contrasting the Engle Granger method, Johansen (1988) co-integration technique is more genuine. The elementary condition for execution Johansen (1988), Johansen and Juselius (1990) is the stationary of variables at first difference. Using this technique, Granger causality test is performed. Equation (20) illustrates this:  t ¼ o1

n X

iXt  i þ

n X

i¼1

Xt ¼ o2

n X

biYt  j þ e1t:

ð20Þ

j¼1

kiXt  i þ

i¼1

n X

diYt  j þ e2t:

ð21Þ

j¼1

7 Econometric model results The results are given below: 7.1 Unit root results The results of the unit root test ADF explain that both the variables economic growth rate and terrorism are not stationary at level, but after taking the first difference of both the variables, they become stationary (see Tables 1, 2). 7.2 Results of granger causality Null hypothesis

Obs

GROWTH does not Granger cause TERRORISM

36

TERRORISM does not Granger cause GROWTH

Table 1 Economic growth rates and economic desgrowth rates of Pakistan (2000–2013)

S economic desgrowth, P terrorism frequency rate Source economic survey of Pakistan

123

Year

Economic growth rate

F-statistic

Prob.

0.04541

0.8326

6.19054

0.0181

S

P

2000

2

-0.464

0.62

2001

3.1

-0.01805

0.44

2002

4.7

-0.02889

0.54

2003

7.5

-0.02874

0.70

2004

9

-0.0937

0.64

2005

5.8

-0.02229

0.37

2006

6.8

-0.10125

0.61

2007

3.7

-0.45271

0.58

2008

1.7

-0.36147

0.47

2009

3.1

-1.02414

0.68

2010

3

2011

3.7

-.81179

0.74

-1.69174

0.79

2012

4.4

-1.98429

0.76

2013

3.6

-1.71351

0.72

How terrorism affects the economic performance? Table 2 Unit root results (ADF)

a

1 % level of significance

Variables

t statistic -3.51

Probability

Level

Growth rate

Level

Terrorism

.0146

First difference

Growth rate

-5.08

0.0003

First difference

Terrorism

-5.44a

0.0001

0.505

.0.98

Table 3 Johansen co-integration tests Null

Alternative

Max statistics

Critical values 5%

r=0

r=0

rB1

r=1

20.56

Trace statistics a

Prob

14.26

0.004

3.84

0.90

0.013

Critical values Proba

5% 20.57

15.49

0.007

3.84

0.07

0.013

Trace test and Max-Eigen value test indicates 1 co-integration equations at the 0.05 level a

MacKinnon-Haug-Michelis (1999) p values

Results of the granger causality confirm that terrorism can cause economic growth but economic growth does not cause terrorism in case of Pakistan. This granger causality results support our model results of ECM-Model. This is concluded from the results that terrorism can affects the economic performance.

8 Results of Johansen co-integrations Results of Johansen co-integration test shows that there is at least one co-integration vector at 5 % level of significance. This confirms that there is long run relationship between terrorism and economic growth (see Table 3). The long run coefficient value of the parameters as below. Economic growth ¼ 5:93  :007 terrorism

ð22Þ

These long run co-efficient values explain that there is a long-run negative and significant relationship between terrorism and economic growth (the value in parenthesis is t-statistic). Again these results support our ECM-Model analysis. 9 Conclusion and policy recommendations Although number of research studies portrayed the negative impact of terrorism on economic growth, but none of the studies depicted its precise magnitude. This study therefore attempts to bridge up this gap. By using ECM-Model, the study quantified the impact of terrorism attacks on economic growth of Pakistan. In this way four major indicators were computed. First, total terrorism frequency rate; second, Pakistan’s terrorism vulnerability rate; third is terrorism devastation magnitude rate and fourth, terrorism vulnerability surface. Furthermore, the economic desgrowth is calculated from the terrorism devastation magnitude rate and terrorism vulnerability rate results.

123

A. Khan et al.

The results of the application of ECM-Model show that the Pakistan vulnerability rate ranges from minimum 0.2 to maximum 0.94. In the selected time period under study, the terrorism devastation magnitude rate is highest in 2013 with value of -2.62 % and lowest in year 2000 with value of -0.09 %. Likewise, impact of terrorism on economic growth is lesser in years 2000 (-.0.04 %) and gradually increased with the highest value of -1.71 % in 2013. It implies that terrorism has caused 1.71 % desgrowth in the Pakistani economy. Keeping in view Pakistani economy, these impacts are substantial and have offset the economic performance of Pakistan. In a nutshell, results of the study conclude that terrorism has created economic disaster in Pakistan. It has hampered economic growth by devastating human and physical capital. These devastations show the leakages from economic growth and consequently the cause of poor economic performance. In this backdrop, this study provides useful policy insights by quantifying the impact of terrorism on economic performance. It will guide policy makers to properly allocate the resources, because if terrorism impacts are underestimated, then it may lead towards misallocation of resources. The government may allocate less budget for defence and security measures, which further may deteriorate the law and order situation. While on the other side, if the impacts of terrorism are overestimated, it may lead to over-allocation of resources causing to create deficits for developmental funds. This study is first step to measure the impact of terrorism attacks on econo4mic performance of developing countries. By using the methodology of this study, the research can be conducted in other countries where terrorism has created problem.

References Abel, A.B., Bernanke, B.S., Croushore, D.: Macroeconomics, 6th edn. Pearson Education, Boston (2008) Blomberg, S.B., Hess, G.D.: How much does violence tax trade? Rev. Econ. Stat. 88(4), 599–612 (2006) Blomberg, S.B., Hess, G.D., Orphanides, A.: The macroeconomic consequences of terrorism. J. Monet. Econ. 51(5), 1007–1032 (2004a) Blomberg, S.B., Hess, G.D., Weerapana, A.: Economic conditions and terrorism. Eur. J. Polit. Econ. 20(2), 463–478 (2004b) Buckelew, A.H.: Terrorism and the American Response Mira. Academic Press, San Rafael (1984) Dickey, D.A., Fuller, W.A.: Distribution of the estimators for autoregressive time series with a unit root. J. Am. Stat. Assoc 74(366a), 427–431 (1979) Dickey, D.A., Fuller, W.A.: Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica 49, 1057–1072 (1981) Engle, R., Granger, C.W.J.: Co-integration and error-correction: representation, estimation, and testing. Econometrica 55, 251–276 (1987) FBI: What we investigate. http://www.fbi.gov/Albuquerque/about-us/what-we-investigate/priorities (2012). Accessed 15 April 2012 Global Terrorism Database. http://www.start.umd.edu/gtd/. (2013) Government of Pakistan (various issues): Economic survey of Pakistan. Ministry of Finance, Islamabad (2013) Gries, T., Krieger, T., Meierrieks, D.: Causal linkages between domestic terrorism and economic growth. Def. Peace Econ. 22(5), 493–508 (2011) Grossman, H.I.: A general equilibrium model of insurrections. Am. Econ. Rev. 81, 912–921 (1991) Gupta, S., Clements, B., Bhattacharya, R., Chakravarti, S.: Fiscal consequences of armed conflict and terrorism in low-and middle-income countries. Eur. J. Polit. Econ. 20(2), 403–421 (2004) Hsiao, C.: Autoregressive modelling of Canadian money and income data. J. Am. Stat. Assoc. 74(367), 553–560 (1979) Hsiao, C.: Autoregressive modelling and causal ordering of economic variables. J. Econ. Dyn. Control 4, 243–259 (1982) Johansen, S.: Statistical analysis of co-integration vectors. J. Econ. Dyn. Control 12(2), 231–254 (1988)

123

How terrorism affects the economic performance? Johansen, S., Juselius, K.: Maximum likelihood estimation and inference on co-integration with applications to the demand for money. Oxf. Bull. Econ. Stat. 52(2), 169–210 (1990) Keynes, J.M.: The Economic Consequences of the Peace. Macmillan, London (1919) Meierrieks, D., Gries, T.: Economic performance and terrorist activity in Latin America. Def. Peace Econ. 23(5), 447–470 (2012) Nitsch, V., Schumacher, D.: Terrorism and international trade: an empirical investigation. Eur. J. Polit. Econ. 20(2), 423–433 (2004) ¨ cal, N., Yildirim, J.: Regional effects of terrorism on economic growth in Turkey: a geographically O weighted regression approach. J. Peace Res. 47(4), 477–489 (2010) Pigou, A.C.: The Political Economy of War. MacMillan, London (1940) Robbins, L.: The Economic Causes of War. Jonathan Cape, London (1942) Ruiz Estrada, M.A., Ndoma, I.: How crime affects economic performance: the case of Guatemala. J. Policy. Model. 36, 867–882 (2014) Ruiz Estrada, M.A., Yap, S.F., Park, D.: ‘‘Applying the natural disasters vulnerability evaluation model to the March 2011 north-east Japan earthquake and tsunami’’. Disasters 38(s2), s206–s229 (2014) Sandler, T., Enders, W.: Economic consequences of terrorism in developed and developing countries. Terror. Econ. Dev. Polit. Openness 17, (2008) Sultan, M.: Terrorism and the macroeconomy: evidence from Pakistan. Def. Peace Econ. (ahead-of-print), 1–26 (2013) Shahbaz, M., Shabbir, M.S., Malik, M.N., Wolters, M.E.: An analysis of a causal relationship between economic growth and terrorism in Pakistan. Econ. Model. 35, 21–29 (2013) Tornell, A.: Reform from within. NBER Working Paper No. 6497, National Bureau of Economic Research, Cambridge, MA (1998) Walkenhorst, P., Dihel, N.: Trade impacts of the terrorist attacks of 11 September 2001: a quantitative assessment. Paper presented at the workshop on the economic consequences of global terrorism, Berlin (2002) Weiner, T.: Afghan Taliban Camps Were Built by NATO. The New York Times, Washington, DC (1998)

123