Valuing the Recreational Uses of Pakistan’s Wetlands: An ...€¦ · Valuing the Recreational Uses of Pakistan’s Wetlands 136 up from home and thus incurring travel costs before

NUST JOURNAL OF SOCIAL SCIENCES AND HUMANITIES

Vol.3 No.2 (July-December 2017) pp. 132-151

Valuing the Recreational Uses of Pakistan’s Wetlands:

An Application of the Travel Cost Method

Iftikhar Hussain Adil* and Ali Dehlavi†

Abstract:

The Keenjhar Lake, Pakistan’s largest freshwater lake and a Ramsar site, is

located in the Lower Indus Basin of the Indus Ecoregion. Global 200, which

scientifically ranks outstanding terrestrial and aquatic ecosystems in 238 ecoregions

worldwide, the Indus Ecoregion is one of the 40 priority Ecoregions. This study uses a

single-site truncated count data travel cost method to estimate the access values of

visitors to Keenjhar Lake. Policy makers may use these estimates on the recreational

value of the lake to assess the returns on conservation investments. A basic version of

the model applied to a subset of visitors using charter transportation allows analysis

of impacts on welfare measurement from altering assumptions about embarkation

points. This study finds the assumption that this category of visitor does not incur travel

and time costs before boarding charter transport to be both unrealistic and simplifying,

leading in turn to an underestimate of consumer surplus values. The strongest argument

in favour of revising data collection and processing strategies in this regard is perhaps

the finding that shared and rented transportation is common in developing countries,

while cost coefficients tend to figure prominently in welfare measurement irrespective

of the functional form.

Keywords: Travel Cost Method, Truncated Count Data Model, Ecotourism, Keenjhar

Lake

1. INTRODUCTION

Keenjhar is Pakistan’s largest freshwater lake (14,000 ha) and is

situated approximately 120km north of Karachi. A wildlife sanctuary

and a Ramsar site, it is set in a stony desert composed of alternating

layers of sandstone and limestone. Approximately 50,000 people, from

12 large and 20 small surrounding villages, are dependent on the lake.

Another predominant use of the lake, which might be labeled indirect

* Iftikhar Hussain Adil <[email protected]> is Professor of Economics at

School of Social Sciences and Humanities (S3H), National University of Sciences and

Technology, Islamabad, Pakistan. † Ali Dehlavi <[email protected]> is Advance Data Analytic Lead, Data Strategy

& Analytic Unit HBL, I.I. Chudrigar Road, Karachi, Pakistan.

133 Adil and Dehalvi

because consumption occurs off-site, is the supply of water for

residential and commercial use in Karachi. The major, direct

consumptive use of the lake among the local population takes the form

of fishing. However, tourists, mainly from Karachi, also enjoy swim-

ming, boating, and other entertainment activities offered by the Sindh

Tourism Development Corporation (STDC) at a resort on the lake’s

western banks.

This study estimates the access value of Keenjhar using a travel

cost method (TCM). It is expected that this would replace existing

decision-making with regard to pricing which does not rely on

quantitative tools but on intuition and experience. Adding the use value

of recreation to the already measured use value of fisheries and other

indirect use values such as the water supply by determining the need to

preserve Keenjhar would provide the planners with more accurate

estimates of its value. After reviewing the existing literature, we have

confined the modeling approach to a count data model for a single site.

The study also addresses the issues associated with multiple purpose

trips and the impacts of labour decisions on time valuation, in addition

to truncation and endogenous stratification via data analysis.

The total economic value of Keenjhar Lake, based on a recent

estimate of the direct consumptive use value (i.e., the producer surplus

from commercial fisheries), the indirect use value (i.e., the residential

water supply to 1 million of the 15 million population of Karachi), and

the non-use value (based on an application of the “choice experiment”

technique administered in Karachi to examine the willingness to pay for

species protection) is in the order of PKR 9 billion annually or USD

145million [Dehlavi, et al. (2008)].

At present, STDC does not employ valuation or similar advanced

quantitative techniques in their planning or pricing of accommodation

and recreational activities. This is unfortunate as models of recreational

demand can be put to a number of uses, including addressing economic

(for e.g., measuring the welfare derived from the reserve) as well as

financial (for e.g., responsiveness to cost components with bearing on

overall revenue or revenue per unit of on-site paying activities)

questions. This paper addresses the economic question of whether

investments in recreational sites provide a return on equity by estimating

Valuing the Recreational Uses of Pakistan’s Wetlands 134

the monetary access value figure associated with the recreational uses of

the Lake.

This study addresses labour market constraints while estimating

time costs by distinguishing between recreationists who are committed

to a fixed work week and fixed vacation allotments and those who are

not constrained in this fashion. The approach followed in this study is

developed by Bockstael, Strand and Hanemann (1987) which argued that

discontinuous labour market constraints lead to corner or interior solu-

tions.

In the case of Keenjhar, it is necessary to take into account the

concerns of the public regarding polluted water which is not only to its

recreational use but also to domestic and commercial uses of the Lake

by Karachi and a local population, mainly inhabitants of the surrounding

twelve large and twenty small villages [WWF-Pakistan (2006)].

Amongst others, pollution of the lake is caused by upstream tanneries,

sewerage, and grease from vehicle-washing and motorized fishing boats.

In a noteworthy economic and epidemiological contingent valuation

survey undertaken at two beaches, Georgiou, et al. (1996) established

that the British public was prepared to pay an amount in excess of the

total clean-up cost that would be incurred to bring British beaches up to

the standard required by the European Community (which in 1995 was

approximately GBP 9 billion).

TCM is used to a subset of visitors using charter transportation.

This was possible owing to a design feature in the questionnaire that

permitted us to analyze the impacts on welfare measurement of visitors

with different embarkation points for their trip to the Lake. The

unrealistic and simplifying assumption that this category of visitor does

not incur travel and time costs before boarding charter transport results

in an underestimate of consumer surplus values. We understand that our

analysis makes the case for revising data collection and processing

strategies since shared and rented transportation is common in

developing countries while cost coefficients tend to figure prominently

in welfare measurement irrespective of the functional form. This study

is useful in Pakistan to estimate non-market values for public policy

purposes. An earlier study by Khan (2004) adopted the TCM with an

objective to shape national policy on the regulation of a national park in


Islamabad while this study is an attempt to shape national policy for

wetlands.

2. THE STUDY SITE AND SAMPLING

Choice of the study site in part was motivated by the STDC’s

own interest in providing economic values for the recreational services

it provides. However, in addition to aiding management decisions, we

were also interested in complementing Keenjhar’s total economic value

estimates USD 145 m [Dehlavi, et al. (2008)].This includes a direct use

component for commercial fisheries, which would be complemented by

recreational use, another direct use value. Even if an augmented value

were not to be applied in a benefit cost analysis in the context of Keenjhar

since it already benefits from protected status as a Ramsar site, the

exercise would be useful for its replicability within the Indus Ecoregion

and elsewhere in Pakistan. Furthermore, the valuable dataset accomp-

anying our analysis allows for further analysis, which would help with

the allocation of scarce resources, allowing planners in turn to assess

whether and by how much to raise fees depending on the magnitude of

the consumer surplus.

Seven-day reconnaissance survey (in February and March of

2009) was held for the purpose of designing a reliable survey instrument.

We conducted a count at the two entrance gates of the site which showed

that 5,892 individuals had visited it during this period. Based on the

findings from the reconnaissance survey, an innovative survey

instrument design element was generated that was incorporated into the

final questionnaire: that is, questions which partitioned travel costs for

those using chartered transport into within-city travel to a common point

of departure and travel onwards to Keenjhar. The form of chartered

transport mentioned here refers to the renting of a bus/van typically by a

single but large family. As there was no reason to assume that all

members of an extended family were picked up from their front door, we

asked respondents using chartered transport if they incurred time and

petrol costs to reach a “common point of departure”. During the main

survey (undertaken between 12-18 August, 2009), the chartered mode

remained the most popular (59 percent), with only a fraction not picked


up from home and thus incurring travel costs before boarding the

chartered transport. Privately owned cars (35 percent) and motorcycles

(6 percent) came second and third among preferred modes of travel.

Our Sampling plan for 1,000 observations which was designed

based on weights of the total observed participation in: (a) activities on

zone basis; (b) activities by each day of a 7-day week; (c) activities by

time periods within a single day and (d) activities by category. This

formulation yielded a convenient way to determine the specific number

of questionnaires to be filled within a given zone, day, time, and activity

category.

The main survey was conducted from 12th to 18thAugust, 2009,

and coincided with a national holiday, the Pakistan Independence Day,

which fell on a Friday in 2009. The survey yielded a sample of 741

visitors.

3. METHODOLOGY

The paper estimates access value to Keenjhar using the TCM.

After describing the theoretical construction of TCM, welfare

measurement using Poisson and negative binomial regression models is

described. This section described the analytical techniques used to

address the separate issues of multiple purpose trips, the impact of labour

decisions on time valuation, truncation and endogenous stratification.

3.1. The Model

The basic recreational demand model used in this paper may be

written as follows:

Pr(𝑥𝑗 = 𝑛) = 𝑓 (𝑛, 𝑧𝑗 , 𝛽), 𝑛 = 0, 1, 2, . . . , 𝑘 … (1)

where, the demand variable x can take an integer value from 0 to k; 𝑧𝒊 is

the row vector of M demand arguments (including the vector of prices

and qualities for recreational sites and the amount of income that could

be earned if the person worked all of the available time); and, β is an M

1 column vector of parameters to be estimated.


TCMs are based on an idea first introduced by Hotelling (1949).

Researchers can derive resource values through the use of TCM by

estimating a demand curve for complementary market goods (for

example a day visitor’s costs of travelling to Keenjhar) and calculating

the welfare value for the household by integrating between the present

price faced by the household for the complementary good and the choke

price, i.e., the price at which the quantity demanded goes down to zero.

3.2. Welfare Measurement in the Poisson Regression Model

Welfare measurement or the value of access to Keenjhar in its

general form is calculated and approximated as the willingness to pay for

use of the site. The computation then is that of the area under the utility-

constant demand curve for the site, or the income-constant demand

curve, given expected low income effects and budget shares of recrea-

tional demand models [Haab, and McConnell (2002)]

𝑊𝑇𝑃(𝑎𝑐𝑐𝑒𝑠𝑠) = ∫ 𝑓(𝑠, 𝐶2𝑖 + 𝑤𝑖𝑡2𝑖𝑝∗

𝑝𝑖0 , 𝑦𝑖

𝑓)𝑑𝑠 … (2)

where, iiii twcP 11

0 (P, the price of a trip to the primary site) and P*is

the relevant choke price; ( iii twc 22 denotes travel cost to the substitute

site, c denotes the round-trip travel cost, w is the after-tax wage rate, and

t is a unit of time for the trip, while 𝑦𝑖𝑓 is a measure of full income, i.e.,

the amount that would be earned if all available time were used up for

work, and s is the dummy variable of integration). Each household is de-

noted by i, while subscripts 1 and 2 index the primary and substitute sites.

As this is an on-site sample with the number of visits expressed as counts,

Poisson regression model is suitable whose probability density function

is given by Haab and McConnell (2002) as follows

Pr(𝑥𝑖 = 𝑛) = 𝑒−𝜆𝑖𝜆𝑖

𝑛

𝑛! 𝑤ℎ𝑒𝑟𝑒, 𝑛 = 0, 1, 2, … … (3)

where, λi > 0 is mean of number of visits and according to characteristics

of poisson distribution equal to the variance.


𝜆𝑖 = exp(𝑧𝑖, 𝛽) … (4)

Willingness to pay for access using the Poisson regression model

is calculated and assuming an exponential function, the choke price is

infinite. Defining P0 as the current travel cost, consumer surplus for

access is given by Haab and McConnell (2002):

𝑊𝑇𝑃(𝑎𝑐𝑐𝑒𝑠𝑠) = ∫ 𝑒𝛽0+𝛽1𝑠𝑑𝑠∞

𝑃𝑖0 = [

𝑒𝛽0+𝛽1𝑠

𝛽1]

𝑃𝑖0

𝑃⟶∞

= −𝑥

𝛽1 … (5)

where, β1 < 0

The Poisson regression model is commonly used in recreational

demand models. Since the Poisson regression model is subject to

misspecification owing to its implicit restriction on the number of

counts: 𝐸 (𝑥𝑖|𝑧𝑖, 𝛽) = 𝑉 (𝑥𝑖|𝑧𝑖 , 𝛽) = 𝜆𝑖 (the conditional mean and

variance are equal). One consequence of variance exceeding the mean

(over dispersion), as is characteristic in recreational data, is that the

Poisson regression model’s standard errors are underestimated, leading

often to the rejection of the null hypothesis of no association. The

Negative Binomial is used to test for over dispersion, a common version

of which is a Poisson regression model with a gamma distributed error

term [Green (2005)]. In such a case, the Negative Binomial’s probability

function can be written as Haab and McConnell (2002):

Pr(xi) = Γ (xi+

1

α)

Γ (xi+ 1) Γ (1

α)

(1

α1

α+ 𝜆𝑖

)

1

α

(𝜆𝑖

1

α+ 𝜆𝑖

)

xi

… (6)

where, 𝜆𝑖 = exp(𝑧𝑖, 𝛽)

The mean and variance of the negative binomial distribution are

𝐸 (𝑥𝑖) = 𝜆𝑖 = exp(𝑧𝑖, 𝛽) and 𝑉 (𝑥𝑖) = 𝜆𝑖 (1 + 𝛼𝜆𝑖), respectively.

The parameter is the over dispersion parameter. If α > 0, over dispersion

is said to exist. If α = 0, no over dispersion or under dispersion exists and

the Negative Binomial collapses to the Poisson distribution in the limit.

If, on the other hand, α < 0, the data are under dispersed so that the

Poisson regression model should be rejected in favour of the Negative


Binomial model, revealing that the test is also one of the Negative

Binomial models against the null hypothesis of a Poisson.

3.3. Endogenous Stratification and Truncation

Count models with truncated samples, i.e., models where only

those visiting the site are sampled, should make use of the appropriate

functional form and also be observant of the effects of functional form

choice and truncation on consumer surplus estimates [Ozuna, et al.

(1993)].

It is expected the sample average number of trips to be higher

than the population mean (endogenous stratification) since on-site

interviewing process is inherently likely to have intercepted avid visitors

to Keenjhar. To obtain the correct likelihood function, it is needed to

account for this oversampling of visitors who have a high use level.

Endogenously stratified and truncated Poisson Regression Model may

be estimated by running a standard Poisson regression of 𝑥𝑖 − 1 on the

independent variables and can be written as its probability [Haab and

McConnell (2002)]:

ℎ (𝑥𝑖 𝑎𝑛𝑑 𝑖𝑛𝑡𝑒𝑟𝑣𝑖𝑒𝑤|𝑥𝑖 > 0) = 𝑒𝜆𝑖𝜆

𝑖

𝑤𝑖

𝑤𝑖! … (7)

where, 𝑤𝑖 = 𝑥𝑖 − 1 and right hand term is the probability function for a

Poisson distribution for the random variable wi. To address over

dispersion relative to the Poisson, truncation at zero, and endogenous

stratification due to oversampling of frequent visitors at Keenjhar,

Negative Binomial regression model should be estimated [Haab and

McConnell (2002)].

𝒉(𝒙𝒊𝒂𝒏𝒅 𝒊𝒏𝒕𝒆𝒓𝒗𝒊𝒆𝒘|𝒙𝒊 > 𝟎) = 𝐱𝐢𝚪 (𝐱𝐢+

𝟏

𝛂)

𝚪 (𝐱𝐢+ 𝟏) 𝚪 (𝟏

𝛂)

(𝟏

𝛂𝟏

𝛂+ 𝝀𝒊

)

𝟏

𝛂

(𝝀𝒊

𝟏

𝛂+ 𝝀𝒊

) 𝛌𝐢𝐱𝐢−𝟏 … (8)


3.4. Multiple Purpose Visits

The model, as is standard, distinguishes single purpose visits

from multiple purpose visits, that is, visits made to destinations on the

way to Keenjhar or on the way back home. This turned out to be very

important since as much as 42 percent of our sample undertook

incidental visits. Using a somewhat recent approach Parsons and Wilson

(1997), one can interact a dummy variable with price to capture both the

shift and rotation of the demand function due to the existence of

complementary sites, thereby adjusting the reported total trip cost of

multiple purpose visitors in our sample.

3.5. Implications of Labour Decisions on Time Valuation

Model in the study also attempted to reflect the implications of

labour decisions on time valuation (opportunity cost of time) and

allowed these decisions to vary over individuals in our sample. In

particular, adopting an approach based on Bockstael (1987), this study

distinguished visitors to Keenjhar who give up on the opportunity to earn

income for a day trip to the Lake from those who do not face any such

trade off. The former “unconstrained” category is different from the

“constrained” one in that it describes individuals whose labour/leisure

choice is at an “interior” and whose opportunity cost of time is reflected

in the wage rate. The modeling structure adopted here is:

𝑥𝑖 = ℎ𝐼(𝑃𝑖 + 𝑤𝐷𝑡𝑖, 𝑃𝑜 + 𝑤𝐷𝑡𝑜 , �̅� + 𝑤𝐷�̅�)

𝑥𝑖 = ℎ𝐶(𝑃𝑖, 𝑡𝑖, 𝑃𝑜 , 𝑡𝑜 , �̅�, �̅�)

𝑙𝑛𝑥𝑚1 = 𝑔 (𝑃𝑚 + 𝑊 𝑡𝑚, 𝐼 + 𝑊 𝑇) 𝑓𝑜𝑟 𝑚 = 𝑗, 𝑘, 𝑙 … (9)

𝑙𝑛𝑥𝑚2 = 𝑔 (𝑝𝑚, 𝑡𝑚, 𝐼 , 𝑇) 𝑓𝑜𝑟 𝑚 = 𝑗, 𝑘, 𝑙 … (10)

where, Equation [9] describes the natural log of the quantity of trips

demanded in the mth mode by unconstrained individuals and Equation 10

describes this with reference to constrained individuals; pm is the travel

cost and tm is the travel time, both associated with the mth mode; w is the

wage rate, I is household income, and T is discretionary time. Time and

budget constraints are collapsed into a single constraint in the case of


Equation (9), which describes “unconstrained” individuals, while time

and budget constraints are separately binding in the case of Equation (10)

“constrained” individuals.

4. RESULTS AND DISCUSSION

The study began by selecting the best estimator for our TCM

through estimating two simple versions of our model. One version uses

travel cost and income variables while the other adds travel time as a

separate variable for so-called “constrained” individuals. Table 1 present

the descriptive statistics of the explanatory variables and theoretically

expected signs. While in Table 2, the two estimators tested were the

Poisson and Negative Binomial, both corrected for zero-truncation and

endogenous stratification. As over dispersion is characteristic in

recreational data, the Negative Binomial tested for over dispersion but

found Poisson more suitable as compare to Negative Binomial.

Table 1. Explanatory Variables and Associated Hypotheses

Variable +/- Mean Std. Dev. Min Max

Travel Cost (TC) (PKR) - 1,283 1,162.30 0 16,078

Travel Time (Ti) (minutes) - 174.05 90.83 3 900

Household Income (mon_income

PKR) + 43,000 74,343.36 2,000 1,000,000

Education (Years of schooling) + 11.85 3.58 0 21

Age (Years) - 31.66 9.97 12 73

Distance (KM) - 49.57 79.84 28 487

Interacted_TC ? 568.29 969.97 0 8212

Interacted_Travel_Time

Gender (male=1, Female=0)

Unmarried (yes=1, No=0)

Region(Urban=0, Rural=1)

?

+

+/-

-

1.65

0.99

0.43

0.85

2.95

0.11

0.50

0.36

0

0

0

0

30

1

1

1

Source: Survey (12-18 August 2009); the sample size is 741.

Having determined an appropriate model, the endogenously

stratified and truncated Poisson alone to estimate demand models were

selected, the first time using seven variables and the next time eleven. In


each of the seven- and eleven-variable versions of the TCM, we carried

out one regression in which incidental consumption effects are ignored

(Model 1); one in which an indicator variable for multiple destination

visits is included (Model 2); and a third one, a fully interacted model, in

which the indicator variable is interacted with the price variable and

travel time (Model 3).

4.1. Estimator Selection for the TCM

The study began by noting some interesting results from the

estimator selection process before proceeding to discuss the travel cost

model. Firstly, in the case of the endogenously stratified and truncated

Poisson, the signs on our coefficients were as expected (see, Table 1).

Secondly, for the purpose of computing the marginal change with

variables held at their means, it is possible to interpret the coefficient of

the total trip cost as semi-elasticity showing the percentage by which

visits would drop for a single unit (i.e., PKR 1) increase in TC. The

coefficients show an extremely low degree of elasticity in both models.

As discussed in Section 3 above, a Keenjhar trip is a unique and high

quality recreational experience and, based on the empirical evidence

reviewed in Woodall, et al. (2002), unique and high quality recreational

experiences have been shown to have low price elasticities. In an

analysis of day trips to Canyon County for wine tourism, the same

authors include a “stay home” dummy variable which confirms the site

in question to be unique, with few substitutes.

As it is an on-site sample with self-selection on the part of those

who go to Keenjhar, the data are truncated because only those visitors

are considered who have visited Keenjhar Lake at least once. Hence,

zero-truncated Poisson regression model is taken. Again, owing to the

fact that ours is an on-site sample, it is inherently likely to have

intercepted avid visitors to Keenjhar. The endogenous stratified

truncated Poisson addresses this problem because it is expected that the

sample average number of trips to be higher than the population mean.


4.2. Estimation of the TCM Model

The results for the selected estimator; the zero-truncated and

endogenously stratified Poisson regression model are presented in

Tables 3 and 4. Seven-variable model is discussed first. Its overall

significance, as measured by χ2, was 0.0000 in all three models showing

that the sign of the coefficients across all variables in Table 3 is as

expected. In general, when displaying marginal effects for travel cost

models we are presented with the predicted number of events in the

dependent variable against each of the independent variables.

Table 2. Estimator Selection for the Travel Cost Model

Endogenous Stratified and

Truncated Poisson

Endogenous Stratified

Negative Binomial

1 2 3 4

Variable Estimate

(S.E.)

Estimate

(S.E.)

Estimate

(S.E.)

Estimate

(S.E.)

Constant

-0.195***

(0.053)

0.373***

(0.065)

-14.347

(196.956)

-12.988

(93.716)

Travel Cost -0.00005*

(0.00003)

-0.00006**

(0.00004)

-0.00005

(0.00005)

-0.00005

(0.00005)

Monthly

Income

1.09 e-08

(5.13 e-07)

7.58 e-08

(5.09 e-07)

-3.62 e-08

(7.45 e-07)

-8.29 e-09

(7.46 e-07)

Travel Time -0.048***

(0.0106)

-0.051***

(0.016)

LR / Wald χ2 2.41 24.12 1.16 11.84

Level of sig. 0.2998 0.0000 0.5601 0.0080

Pseudo R2 0.0010 0.0096

Note: ***, ** and * indicate significance at the 1 percent, 5 percent and 10 percent levels, respectively.

Results are for a sample size of 741.

Based on a high Pseudo R2 value as compared to Models 1 and

2, the marginal effects in Model 3 are examined in Table 3. It can be

observed that trips are predicted by the Model to increase by 0.03 percent

for a PKR. 1000 increase in monthly income. By looking at elasticities

after Poisson, a 10 percent increase in travel costs would result in a 1.3

percent decrease in the frequency of trips. It is noteworthy that travel

cost variable is significant at the 5 percent level for all three models.


The coefficient of monthly income is insignificant in all the

models, considering that nearly half the sample has undertaken an

approximately 3-hour long journey. Supporting hypotheses that males,

more importantly single males, face nominal constraints when it comes

to traveling unaccompanied and exercising travel decision prerogatives,

gender and married coefficients show significance at the 5 percent and

10 percent levels, respectively.

Table 3. Endogenous Stratified and Truncated Poisson

Regression- Basic Models

1 2 3

Variable Estimate

(S.E.)

Estimate

(S.E.)

Estimate

(S.E.)

Constant

-1.192

(0.728)

-1.288*

(0.729)

-1.218*

(0.729)

Travel Cost -0.00006**

(0.00003)

-0.00007**

(0.00004)

-0.0001**

(0.00004)


(0.011)

-0.053***

(0.011)

-0.080***

(0.015)

Monthly Income 2.17 e-07

(5.06 e-07)

2.91 e-07

(5.11 e-07)

3.19 e-07

(5.13 e-07)

Gender 1.696**

(0.709)

1.719**

(0.709)

1.783**

(0.710)

Married 0.231***

(0.072)

-0.216***

(0.072)

-0.218***

(0.072)

Urban 0.299***

(0.088)

-0.301***

(0.088)

-0.304***

(0.088)

Waterac_pref 0.255*

(0.139)

0.234*

(0.139)

0.220

(0.139)

D_mp 0.225***

(0.070)

-0.048

(0.132)

Interacted_TC 0.00007

(0.00007)

Interacted_T_Time 0.058***

(0.022)

LR (χ2) 58.19 68.5 76.43

Level of sig 0.0000 0.0000 0.0000

Pseudo R2 0.0233 0.0274 0.0306

Note: Note: ***, ** and * indicate significance at the 1 percent, 5 percent and 10 percent levels, respectively.

The correlation between the decision to travel to Keenjhar and

the preference for water-based activities appears to be supported by the


significance of the coefficient of our “water based activities” variable at

the 10 percent level except in Model 3. Moreover, the coefficient of the

dummy variable for multiple purpose visits is highly significant in Model

2 but not so when interacted with travel time in Model 3.

Table 4. Endogenous Stratified and Truncated Poisson

Regression - Extended Models

Variable

Model 1 Model 2 Model 3

Estimate

(S.E.)

Estimate

(S.E.)

Estimate

(S.E.)

Constant -1.299*

(0.751)

-1.424*

(0.752)

-1.352*

(0.753)

Travel Cost -0.00007**

(0.00003)

-0.00007**

(0.00004)

-0.0001***

(0.00005)


(0.011)

-0.052***

(0.011)

-0.079***

(0.016)

Monthly Income 1.72 e-07

(5.22e-07)

2.31 e-07

(5.29 e-07)

2.65e-07

(5.32 e-07)

Gender 1.699**

(0.710)

1.725**

(0.711)

1.777**

(0.711)

Married -0.254***

(0.085)

-0.241***

(0.084)

-0.239***

(0.085)

Urban -0.296***

(0.089)

-0.310***

(0.089)

-0.304***

(0.089)

Waterac_pref 0.254*

(0.139)

0.234*

(0.139)

0.221

(0.139)

Education 0.004

(0.010)

0.006

(0.010)

0.007

(0.010)

Unemp_09 0.061

(0.089)

0.073

(0.089)

0.050

(0.090)

Wtp_50 0.080

(0.080)

0.083

(0.079)

0.094

(0.080)

Age -0.0005

(0.004)

-0.0004

(0.004)

-0.0005

(0.004)

D_mp 0.231***

(0.071)

-0.047

(0.133)

Interacted_TC 0.00007

(0.00007)

Interacted_Travel_Time 0.058***

(0.022)

LR χ2 60.10 71.00 78.95

Level of Sig. 0.0000 0.0000 0.0000

Pseudo R2 0.0240 0.0284 0.0316

Note: ***, ** and * indicate significance at the 1 percent, 5 percent and 10 percent levels, respectively.

The mean consumer surplus result relates to Model 1 and needs

to be interpreted in light of Models 2 and 3 (Table 3) which exhibit


incidental visits to complementary sites such as Badshahi Masjid,

Bhamboor, Makli and Chowkandi. In Model 2, following the introduct-

ion of an indicator variable, the statistical significance of the differential

intercept implies that the intercept for the multi-purpose trips (the 58

percent of our sample undertaking single purpose visits and the 42

percent undertaking multiple purpose visits) is different. As the coeffi-

cient is positive, the incidental visits may be interpreted to serve as

complements [Loomis, et al. (2000)].

All the models estimated are presented in Tables 3 and 4 while

segregating visitors to Keenjhar who forego the opportunity to earn

income for the purpose of enjoying a day trip to the Lake from those who

do not. The former “constrained” category describes individuals whose

labour/leisure choice is at an “interior” and whose opportunity cost of

time is reflected in the wage rate. In demand function, the opportunity

cost of time of the latter “unconstrained” category is reflected in the

absolute value of the coefficient of the travel time variable in Tables 3

and 4.

Constrained individuals’ opportunity cost of time was based on

empirical evidence presented by Cesario (1976) where the most common

assumption is that the price of time spent travelling can be valued at

between a ¼ and ½ of the wage rate. The unconstrained individuals’

time was set at zero. When this structure of modeling is not followed,

and all individuals are treated equally for time valuation, the estimated

mean consumer surplus per visit is PKR27,322 (or USD329). As shown

in Table 4, for Model 3, the estimated mean consumer surplus per visit

in the endogenous stratified truncated Poisson is PKR9,024 (or

USD109). This estimate is only marginally smaller (a 5 percent diff-

erence) when compared to the result obtained from our 7-variables

endogenous stratified truncated Poisson regression.

4.3. Impact of Outset Origins on Welfare Measurement

A design feature in our questionnaire permitted the analysis of

impacts on welfare measurement arising from differences among visitors

in terms of their point of departure. In particular, respondents travelling

as a large group in rented buses/vans (437 visitors or 59 percent of our


sample of 741) were asked if they incurred petrol and time costs before

boarding their charter transport. In other words, we did not assume that

all members of an extended family or all members of a group of friends

were picked up from their front door. As referred in Table 5 to a

“common point of departure” (CPD) to describe the point of boarding

chartered transport for those who were not picked up from their home (6

percent of sample). The term “home” indicates that welfare was cal-

culated on the assumption that all chartered transport visitors were

picked up from their doorstep. Common point of departure “CPD” and

“home” is the embarkation points for each of the two sub-samples that

are used to estimate consumer surplus in Table 5. The impact on

consumer surplus, an increase of 41 percent, is pronounced when the

sub-sample is restricted to only those who reported costs before boarding

their chartered transport.

However, the analysis does show that the unrealistic and

simplifying assumption that this category of visitor does not incur travel

and time costs before boarding charter transport results in a significant

underestimate of consumer surplus values. The design feature is cumber-

some and cause respondent fatigue. Moreover, in a review of revealed

preference valuation techniques, Bockstael (2007) found that cost

coefficients tend to figure prominently in welfare measurement

irrespective of the functional form. With regard to the zonal travel cost

method, Bateman, et al. (1997) undertook research on embarkation

points for the trip emphasizing the accuracy of road distances and routing

to recreational sites and the impact that this has on welfare measurement.

This study was based on a sample of 351 visitors to a woodland

recreation site. The authors use actual road network distance in order to

compare the consumer surplus estimates derived using it with those

obtained by assuming straight line travel, where they found the latter to

underestimate welfare values up to 20 percent. The revision of data

collection and processing strategies in our case is all the more important

given the fact that chartered transport of the kind used by Karachiites is

common in the urban centers of developing countries. For example, in

an application of the travel cost method in Bangladesh, Shammin (1999)

found as many as 58 percent of visitors to the Dhaka zoo to use a bus as

compared with 20 percent in the tempo/scooter category. The easy avail-

Table 5. Results of Access Values for Different Specification of Time Cost and Out of

Pocket Expenses in the Travel Cost Variable

Sample Used

Outset

Origin

TC

Coefficient

Standard

Error

t - Value Log

Likelihood

Prob > chi2 Consumer

Surplus

(mean per

visit, USD)

47 charter transport users

who were not picked up

from home (6 percent of

the sample)

CPD -0.000189 0.000374 -0.51 -47.23 0.12 64

Home -0.000266 0.000373 -0.71 -47.10 0.11 45

Entire sub-sample of 437

charter transport visitors

(59 percent of the sample)

CPD -0.000281 0.0001 -2.81 -699.2 0.02 43

Home -0.000296 0.000010 -2.97 -698.7 0.02 41

Full sample (741

visitors)*

Home -0.000111 0.000051 -2.19 -1211 0.00 109

Full sample (741

visitors)**

Home -0.000105 0.000050 -2.11 -1213 0.00 115

Note: The term “home” indicates that welfare was calculated assuming that all chartered transport visitors were picked up from their doorstep; conversely, the

welfare measurement incorporating time and out-of-pocket expenses incurred before boarding chartered transport is denoted by “common point of departure”

(CPD).

14

8 V

alu

ing

the R

ecreatio

na

l Uses o

f Pa

kistan

’s Wetla

nd

s


ability of buses, microbuses and other shared/chartered transport for low

income groups visiting popular public attractions is also underlined in

Mahat and Koirala (2006), which applies the travel cost method to study

visitors to the Jawalakhel Central Zoo of Nepal where this category

represented 80 percent of all transport modes.

5. CONCLUSION AND POLICY IMPLICATIONS

This study used a single-site truncated count data travel cost

method to estimate the access values of visitors to Keenjhar Lake. A

basic version of the model applied to a subset of visitors using charter

transportation allowed to analyze the impacts on welfare measurement

from altering assumptions about embarkation points. Policy makers may

use these estimates on the recreational value of the lake to assess the

returns on conservation investments.

Using the TCM, in this study we estimate the recreational value

of Keenjhar Lake. The estimated value is PKR3.46 billion (or USD42.2

million) calculated on the basis of an annualized PKR9,500 (or USD116)

illustrating consumer surplus per visit value, that is, assuming average

daily visits at 1,000.

From the above results it can be inferred easily that amenity value

of the site can be improved by providing facilities to the visitors. It is

well known fact that amenity sites are scare in Pakistan. People chose to

go for recreational trips if facilities are available at the site. Facilities

may be provided to the visitors when the funds are available. Funds can

be generated through increase in entry ticket, introducing parking fee,

taxing vendors and minor percentage raise in the STDC huts charges.

Sindh Tourism and Development Corporation and WWF Pakistan may

collaborate so that to generate revenue and save the environment at the

same time.


REFERENCES

Bateman, I.J., J.S. Brainard, G.D. Garrod and A.A. Lovett (1997) The

Impact of Journey Origin Specification and Other Assumptions

Upon Travel Cost Estimates of Consumer Surplus: A

Geographical Information Systems Analysis. Discussion Paper

No. 28, Department of Economics and Marketing, Lincoln

University, U.K.

Bockstael, N.E. and K.E. McConnell (2007) Environmental and

Resource Valuation with Revealed Preferences – A Theoretical

Guide to Empirical Models. Bateman (ed.) Economics of Non-

Market Goods and Resources, Vol. 7, Dordrecht, Netherlands.

Bockstael, N.E., I.E. Strand and W.M. Hanemann (1987) Time and the

Recreational Demand Model. American Journal of Agriculture

Economics, 69:2, 293–302.

Cesario, F.J. (1976) Value of Time in Recreational Benefit Studies. Land

Economics 52:32, 41.

Dehlavi, A., B. Groom, B.N. Khan and A. Shahab (2008) Total

Economic Value of Wetland Sites on the Indus River. Report by

the World Wide Fund for Nature – Pakistan, Indus for All

Programme, Karachi.

Georgiou, S., I. Langford, I. Bateman and R.K. Turner (1996) Economic

and Epidemiological Investigation of Coastal Bathing Water

Health Risks. Centre for Social and Economic Research on the

Global Environment (CSERGE), Working Paper PA 96-01,

Norwich: University of East Anglia.

Green, W. (2005) Functional Form and Heterogeneity in Models for

Count Data. Foundations and Trends in Econometrics, 1:2, 743-

744.

Haab, T.C. and K.E. McConnell (2002) Valuing Environmental and

Natural Resources: The Econometrics of Non-Market Valuation,

Northampton, MA, USA: Edward Elgar.

Hotelling, H. (1949) Letter to the Director of the National Park Service

(Letter dated June 18, 1947), in Roy A. Prewitt (ed.). The

Economics of Public Recreation: The Prewitt Report,

Department of the Interior, Washington, D.C.


Khan, H. (2004) ‘Demand for Eco-Tourism: Estimating Recreational

Benefits from the Margalla Hills National Park in Northern

Pakistan’. SANDEE Working Paper No. 50-4, SANDEE,

Kathmandu.

Loomis, J., S. Yorizane, and D. Larson (2000) Testing Significance of

Multi-Destination and Multi-Purpose Trip Effects in a Travel

Cost Method Demand Model for Whale Watching Trips.

Agricultural and Resource Economics Review, 29:2, 183-191.

Mahat, T.J. and M. Koirala (2006) ‘Economic Valuation of the Central

Zoo of Nepal’. Paper Presented at the 9th Biennial Conference of

the International Society for Ecological Economics, 15-18

December, 2006, New Delhi, India.

Ozuna Jr., T., L.L. Jones and Oral Capps, Jr. (1993) Functional Form and

Welfare Measures in Truncated Recreation Demand Models.

American Journal of Agricultural Economics, 75:4, 1030-1035.

Parsons, G.R. and A.J. Wilson (1997) Incidental and Joint Consumption

in Recreational Demand. Agricultural and Resource Economics

Review 26, 107-115.

Shammin, M.R. (1999) ‘Application of the Travel Cost Method–A Case

Study of Environmental Valuation of Dhaka Zoological Garden’.

In Joy E. Hecht (ed.). The Economic Value of the Environment:

Case Studies from South Asia. IUCN–The World Conservation

Union, Gland, Switzerland.

Woodall, S., P. Wandschneider, J. Foltz, and R.G. Taylor (2002)

‘Valuing Idaho Wineries with a Travel Cost Model’. Selected

Papers, Western Agricultural Economics Association, 2002

Annual Meeting, Long Beach, California, USA.

World Wide Fund for Nature-Pakistan (2006) ‘Preliminary

Environmental Baseline Study Report: Keti Bunder, Keenjhar

Lake, Chotiari Reservoir and Pai Forest’. Indus for All Program,

Karachi.



Decomposition Analysis of Energy Consumption in

Pakistan for the Period 1990-2013

Faisal Jamil* and Arbab Shahzad†

Abstract:

The final energy consumption in Pakistan has doubled during the last two

decades. Investigating the factors responsible for changes in energy use is important

for future projections. Decomposition techniques enable us to quantify the contributing

factors in aggregate energy change. This study attempts to investigate the factors

behind the aggregate change in energy consumption over the period 1990-2013 using

Logarithmic Mean Divisia Index (LMDI) decomposition technique. LMDI decomposes

the overall change in energy use into three effects namely, activity, structural, and

intensity effects. Results of the study suggest that observed increase in Pakistan’s

energy consumption is primarily due to the activity and structural effects. The energy

intensity of overall economy has decreased showing an increase in energy efficiency,

though at a decreasing rate over time. The quantification of energy imports based on

projections shows that Pakistan may face serious fiscal challenge by 2025 due to

extremely large energy import bill and possible energy price shocks. There is a need to

put efforts towards reducing the gap between energy supply and demand, diversifying

domestic energy production including increased reliance on renewables, efforts

towards energy and environment conservation, and efficient use of available resources.

Keywords: Energy, Decomposition Analysis, LMDI, Intensity Effect

1. INTRODUCTION

Energy is an essential input affecting the output and the overall

economic welfare. Various past studies examine the relationship

between economic growth and energy consumption and find different

results for different economies [Masih and Masih (1996); Soytas and

Sari (2003); Lee (2007); Apergis and Payne (2010); Jamil and Ahmad

(2010); Belke, et al. (2011); and, Shahbaz, et al. (2012)]. The literature

* Faisal Jamil <[email protected]> is Assistant Professor at School of

Social Sciences and Humanities (S3H), National University of Sciences & Technology

(NUST), Islamabad, Pakistan. † Arbab Shshzad <[email protected]> School of Social Sciences and

Humanities (S3H), National University of Sciences & Technology (NUST), Islamabad,

Pakistan.

mailto:[email protected]


153 Jamil and Shahzad

finds that growth in energy consumption is inevitable in the economic

growth although there is lack of consensus on the direction of causality.

The final energy consumption in Pakistan increased from 21.58

million tons of oil equivalents (MTOE) in 1990 to 40.18 MTOE in 2013

[Energy Yearbook (2014)]. The role of economic growth is considerable

in raising energy consumption. The energy mix and consumer mix for

different energy sources are changing over time making the job of policy

makers and energy planners challenging. The investigation of changing

energy consumption patterns and the factors responsible for this change

is an important research topic. Therefore, this study seeks to quantify the

impact of structural shift of economic activity on total energy demand in

Pakistan so as to have better understanding of the energy use

mechanisms.

Energy decomposition is a widely used method that tracks the

relative contributions of various factors to changes in energy consump-

tion, energy intensity and the environment [Ang and Zhang (2000);

Mairet and Decellas (2009)]. Different decomposition techni-ques have

been used in various past studies such as, Arithmetic Mean Divisia Index

method-1 (AMDI-1), Modified Fisher Ideal Index method, Conventional

Fisher Ideal Index method, Shapley/Sun method and Marshall-

Edgeworth method, Laspeyres, Paasche, Sato-Vartia and Tornqvist [Ang

and Zhang (2000); Liu and Ang (2003); Ang (2004)]. Some studies

provide a comparison of various decomposition methods and show that

log mean divisia index (LMDI) method is robust due to being consistent

in aggregation, residual free, and easy interpretation. These advantages

lead to an extensive use of LMDI in different past studies [Ang and Choi

(1997); Ang and Liu (2001); Ang (2004); Ang (2005); Ediger and Huvaz

(2006); Ma and Stern (2008); Hatzigeorgiou, et al. (2008); Mairet and

Decellas (2009); Zhang, et al. (2011); Balezentis and Strienikiene

(2011); Nasab, et al. (2012); Hasanbeigi and Sathaye (2012); Zhao, et al.

(2012); and Lotz and Pouris (2012)].

LMDI enables to investigate the responsible factors of aggregate

energy consumption change over time by decomposing the overall

change into certain effects including activity effect, structural effect and

intensity effect. The activity effect shows the impact of a change in

energy consumption associated with level of economic activity;

Decomposition Analysis of Energy Consumption 154

structural effect summarizes the impact of composition of energy

consuming activities; while, intensity effect highlights the impact

associated with sectoral use of energy per unit of output. In a nutshell,

literature has established the usefulness of decomposition in energy

related forecasting as it helps to track down factors that are responsible

for change in aggregate energy use. Most of the past studies decompose

either total energy consumption or carry out the analysis for a particular

sector especially the manufacturing sector. Moreover, a range of

decomposition methods are found in the literature, but a consensus about

the best decomposition method is still lacking.

This study applies period-wise decomposition analyses for the

total period as well as decade-wise analysis separately for 1990-2000 and

2000-2013 and also quantifies the cost to the economy of the projected

energy imports by 2025. The economy is broadly classified into

agriculture, industrial and services sectors and energy consumption rises

in all the sectors that contributed to the energy imports of Pakistan. The

results suggest that observed increase in the energy consumption is

primarily attributed to the activity and structural effects. The energy

intensity of overall economy shows a decreasing trend signifying

improvements in energy efficiency over the period. However, the

compensating share of intensity effect was higher during the period

1990-2000 as compared to 2000-2013. The sector-wise analysis shows

that agriculture is responsible for a decline in total energy consumption,

while industrial and services sectors have contributed in energy con-

sumption growth. The quantification of energy imports based on official

projections indicates that the country may face serious balance of

payment challenges by 2025 due to heavy reliance on imported energy.

It suggests diversifying energy mix towards more renewable and

sustainable energy sources.

The rest of the paper is as follows. Section 2 provides an over-

view of energy sector in Pakistan. Section 3 describes the methodology

and the data while Section 4 discusses the results. Section 5 shows the

impact of future energy imports on Pakistan’s external account. Finally,

Section 6 concludes the study.


2. AN OVERVIEW OF PAKISTAN ENERGY SECTOR AND

THE ECONOMY

Total final energy consumption is increasing in Pakistan and is

projected to reach 142 MTOE by 2025 [Pakistan Integrated Energy Plan

(2013)]. The final energy mix is mainly comprised of four sources

including natural gas, oil, hydroelectricity and coal. The share of natural

gas and oil is about 74% in the total mix. Petroleum imports constitute

about 30% of total energy use in Pakistan in 2014 and are increasing as

indigenous natural gas resources are depleting. Energy efficiency is the

major future energy source in the world. It is valid in the case of Pakistan

as the energy intensity calculated by energy use per unit of output has

declined over time, from 9.42 TOE in 1990 to 7.09 TOE and 6.64 TOE

per unit of output respectively in 2000 and 2014. Figure 1 shows the

period-wise changing patterns of Pakistan’s fuel mix [Energy Yearbook

(2000, 2014)].

Total energy consumption in different sectors shows changing

pattern as the share of agriculture is declining whereas, shares of industry

and service sectors have slightly increased (see, Figure 2). A similar

pattern is evident in the sector-wise economic activity shares during

1990-2013 as shown in Figure 3. The economy has experienced struc-

tural shifts especially during the last decade wherein the share of

agriculture in the economy falls and the shares of industry and services

sectors increase. The composition of intra-sector economic activity is

given in the subsequent figures. Figure 4 depicts the changing pattern of

sub-sectoral shares in the agriculture sector. The share of crops in agri-

cultural GDP has fallen from 65 to 42% while, livestock subsector

witnessed a phenomenal rise in its share from 30 to 55% over the period.

This structural pattern clearly indicates that agricultural sector has been

rapidly moving from relatively energy intensive crop sub-sector towards

less energy intensive livestock sub-sector. Figure 5 shows the shares of

sub-sectors in the industrial sector. Manufacturing sector’s share in the

total sectoral output has increased from 68% to 73% over the same time

period. The other two sub-sectors, i.e., construction and utilities’ supply

have witnessed decreasing share in total industrial output.


Figure 1. Share of Fuels in total energy Mix in the period 1990-2014

Source: Hydrocarbon Development Institute of Pakistan (2004 and 2014).

Figure 2. Sectoral share in total energy consumption in Pakistan

1990-2013

Source: Hydrocarbon Development Institute of Pakistan (2004 and 2014).

23.83

32

45

38

48.32

2931.72

14.68 166.43 4.98

10

0

10

20

30

40

50

60

1990 2000 2014

Fuel

Mix

% i

n t

ota

l E

ner

gy

Years

Natural Gas Petroleum Electricity Coal

6.92 2.66 1.64

33.9134.62 35.48

59.15 63.06 62.88

0

20

40

60

80

100

120

1990 2000 2013

Per

cen

tage

Shar

e

Years

Agriculture Industry Services


Figure 3. Sectoral share of Economic Activity 1990-2013

Source: Pakistan Economic Survey (Various Issues)

Figure 4. Sub-sector share in Agriculture Output


25.59 25.92 21.05

23.07 23.3225.25

51.33 50.74 53.7

0

20

40

60

80

100

120

1990 2000 2013

Agriculture Industry Services

6550.7

41.66

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990 2000 2013

% S

hare

in

To

tal

Ag

ricu

ltu

re

Crops livestock Fisheries Forestry


Figure 5. Sub-Sector Share in Industrial Output in Pakistan (%)


Figure 6. Sub-Sector Share in Services Output (Percentage)


2.179.77 9.44

68.4562.92

73.42

16 10.51

8.513.38 16.8

8.64

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990 2000 2013

% S

hare

in

To

tal

Ind

ust

ry

Min & Quar Manuf Construction Elect, Gas, Water Supply

19.76 22.14 18.02

34 34.431.98

4.587.3

8.9411.27 6.2

5.08

14.25 12.1812.37

16.14 17.78 23.61

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990 2000 2013

% S

hare

in

To

tal

Serv

ices

Tra & Com WS & RT Fin & Ins OD P Adm & Def SCS


Figure 6 shows sub-sectoral shares in the services. Finance and

insurance sub-sectors have doubled over the period 1990-2013 whereas,

social and other community services rose from 16% to 24%. The analysis

shows that energy use, composition of energy sources, and sectoral share

in energy consumption is changing. It signifies the importance of investi-

gating energy consumption trends across different sectors to formulate

an optimal energy supply system in the country. This analysis will help

in understanding the energy consumption behaviour vis-à-vis the re-

quirements for infrastructure and future planning for making the desired

energy sources available in sufficient amounts. The analysis will also

guide the policy makers to plan for sustainable energy future in the

country.

3. METHOD OF DECOMPOSITION AND DATA

During the last two decades, several studies decompose the

changes in energy consumption into specific predetermined effects to

capture the factors that may explain the change over a period of time

[Choi, et al. (1995); Ang and Liu (1995); Ang and Lee (1996); Hoekstra

and Bergh (2003); Liu and Ang (2003); Ang, et al. (2004); and Reddy

and Ray (2010); Ullah, et al. (2014)]. Ang and Zhang (2000) and Ang

(2004) provide a survey of literature on several decomposition methods

including LMDI, AMDI-1, Modified Fisher Ideal Index method,

Conventional Fisher Ideal Index method, Shapley/Sun method and

Marshall-Edgeworth method. Past studies distinguish between two

alternative techniques of decomposition, i.e., structural decomposition

analysis (SDA) and index decomposition analysis (IDA). SDA method

is based on input-output coefficients, while IDA is based on aggregate

input-output data. Ang and Lee (1996) extend the energy consumption

approach by replacing it with energy coefficient approach where the

impact of structural and sectoral energy efficiency changes are measured

in terms of coefficients and elasticities.

IDA is useful to decompose the change in aggregate variables in

two or more different components with their shares in the aggregate

change. LMDI is one of the index based techniques of IDA [Ang and

Zhang (2000)]. An LMDI with both multiplicative and additive


techniques is proposed by Ang and Choi (1997). LMDI has been

extensively used for decomposition of energy consumption in different

countries and identified the impact of predefined factors in energy

consumption change Ediger and Huvaz (2006); Ma and Stern (2008);

Hatzigeorgiou, et al. (2008); Mairet and Decellas (2009); Zhang, et al.

(2011); and Zhao, et al. (2012)]. In some of the studies, sectoral effects

representing a transition from agriculture to industrialization explains

increase in energy demand [Lotz and Pouris (2012); Zhao, et al. (2012)],

while in many other studies, activity effect is the main driver of change

in aggregate energy consumption [Balezentis, et al. (2011); Nasab, et al.

(2012)]. Hasabeigi, et al. (2012) incorporate additive LMDI approach

for decomposing energy intensity of Californian industries for the period

1997-2008 and find that structural effect that cause a shift from high

energy intensive industries to low energy intensive industries play an

important role in reduction of energy intensity.

Among different IDA techniques, LMDI is found to be a superior

decomposition method in various past studies Ang and Liu (2001); Ang

(2004); and Ang (2005); Xu and Ang (2014). Ang, et al. (2004)] compare

LMDI results with many other methods including (AMDI)-1,

Conventional and Modified Fisher Ideal Index method, Shapley/Sun

method and Marshall-Edgeworth method, Simple Average Parametric

Divisia Index, Laspeyres Index, and Divisia Index Arithmetic Mean

Function and find that LMDI is preferable on the basis of properties like

theoretical foundation, ease of use, result interpretation and adaptability.

Ang and Liu (2001, 2007) reiterate that LMDI a better decomposition

method with no residual term (an overview of some of the selected

studies on the energy decomposition is given in Annexure). Other

identified features of LMDI include comprehensive formula, error free

decomposition, perfect aggregation and capability of handling zero

values in the data set. Due to these merits, both multiplicative and

additive LMDI are considered superior to other methods. Therefore, we

employed LMDI for decomposing total energy consumption in Pakistan.

Decomposition is a top down approach that has been greatly used

for apportioning the aggregates into different components. This study

aims to investigate the factors that are accountable for change in

aggregate energy consumption in Pakistan for the period 1990-2013


method. The total energy consumption is decomposed using LMDI for

three periods of different lengths, (i) the analysis for overall period that

is, 1990 to 2013; (ii) 1990-2000; and (iii) 2000 to 2013. The first set

analyzes the aggregate changes for the whole time period to get historical

trends with the year 1990 as a base period and 2013 as a current period.

The second and third sets are for the decade-wise analysis, where base

years are 1990 and 2000 respectively. Change in aggregate energy use is

a long term phenomena hence, we conduct a period-wise decomposition

rather than time series decomposition. The analysis classifies the overall

economy into three major sectors including agriculture, industry and

services. Main variables in the data set are the final energy consumption

in Tonnes of Oil Equivalent (TOE) and output level (GDP in million

rupees) for all the three sectors. Sector-wise energy consumption data is

obtained from Hydrocarbon Development Institute of Pakistan

(Unpublished) and collected from Pakistan Energy Yearbook

(Published). The economic activity data are collected from various issues

of Pakistan Economic Survey.

The study employs IDA based method of LMDI proposed by

Ang and Choi (1997) and Ang (2005). LMDI is based on an identity,

where the left-hand-side must equal the right-hand-side making a

decomposition model with no residual. There are two main approaches

in LMDI: (a) ratio decomposition also known as multiplicative

decomposition and (b) difference decomposition also known as additive

decomposition. In this study, the latter is adopted, as the former gives

results in ratio form that are difficult to interpret. The method examines

the impact of pre-defined effects explaining the change in total energy

consumption in absolute form by adding all the factors on right-hand side

of the identity. The overall change is divided into the above mentioned

three effects as defined below:

1. Activity effect refers to the change in aggregate energy

consumption due to change in overall activity level.

2. Structural effect refers to the change due to a change in activity

mix by a sub-sector.

3. Intensity effect refers to the use of energy per million rupees of

output.


Let the total energy change be represented by G, assuming that there

are a number of factors represented by n associated with change in the

aggregate over the period of time with quantitative variables, i.e., x1, x2,

x3 , ..., xn. The subscript i represents sub-sector. The general IDA identity

is given as follows:

∑ 𝐺𝑖 𝑛𝑖=1 = ∑ 𝑛

𝑖=1 𝑥1𝑖, 𝑥2𝑖, … , 𝑥𝑛𝑖 … (1)

The aggregate change from G0 to GT takes the form:

𝐺0 = ∑ 𝑥1𝑖𝑜 , 𝑥2𝑖

𝑜 , … , 𝑥𝑛𝑖𝑜𝑛

𝑖=1 and 𝐺𝑇 = ∑ 𝑥1𝑖𝑇 , 𝑥2𝑖

𝑇 , … , 𝑥𝑛𝑖𝑇𝑛

𝑖=1

where, 0 is the base period and T the current period. Decomposition

through additive technique takes the following form.

∆DTot = GT- G0 = ∆Gx1+ ∆Gx2 + … + ∆Gxn. … (2)

Additive decomposition gives results in absolute form by adding all the

factors on the right-hand-side making them exactly equal to the left-

hand-side of the identity. The general LMDI formula for the kth factor

in Equation (2) is as follows:

∆Gxk= ∑ 𝐿(𝐺𝑖𝑇 𝐺𝑖

0)𝑙𝑛 (𝑥𝑘𝑖𝑇/𝑥𝑘𝑖

0𝑛𝑖=1 ) …(3)

The IDA for the three factors case is given below,

E= ∑ 𝑛𝑖=1 Ei = ∑ 𝑄𝑛

𝑖=1 𝑄𝑖

𝑄

𝐸𝑖

𝑄𝑖= ∑i QSiIi … (4)

where,

E = The total energy consumption of all the sectors in the economy

Q = ∑i Qi is the total economic activity/output level of all sectors in the

economy

Qi is the output level of the ith sector

Si = Qi/Q is the activity share of the ith sector

Ei is the energy consumption of the ith sector


Ii = Ei /Qi is the energy intensity of the ith sector

For additive decomposition, we decompose the difference such as,

∆ETOT= ET-E0= ∆EAct+ ∆EStr+∆EInt … (5)

where,

∆ETOT denotes aggregate change in energy consumption,

ET denotes energy consumption in current year,

E0 denotes energy consumption in base year,

∆EAct denotes change in energy consumption due to activity effect,

∆EStr denotes change in energy consumption due to structural effect,

∆EInt denotes change in energy consumption due to intensity effect.

Three effects on right-hand-side of the Equation (5) are estimated by

employing the following Equations:

∆EAct = ∑i [𝑊𝑖 ∗ 𝑙𝑛 (𝑄𝑇

𝑄0)] … (5a)

∆EStr = ∑i [𝑊𝑖 ∗ 𝑙𝑛 (𝑆𝑖𝑇

𝑆𝑖0)] … (5b)

∆EInt = ∑i [𝑊𝑖 ∗ 𝑙𝑛 (𝐼𝑖𝑇

𝐼𝑖0)] … (5c)

where,

𝑊𝑖 = 𝐸𝑇−𝐸0

𝑙𝑛 𝐸𝑖𝑇− 𝑙𝑛 𝐸𝑖0

To calculate activity effect (∆EAct) for the three sectors, we divide total

current period output by total base period output for all three sectors

(QT/Q0) and take their natural log. The results are multiplied to aggregate

Wi of all the three sectors that gives us final value for activity effect. The

structural effect (∆EStr) is calculated for the three sectors by dividing the

current period output share of the ith sector in total output by base period

output share of the same sector in total output, i.e., (SiT/Si

0). Similarly,

intensity effect (∆Eint) is calculated for the three sectors by dividing the

current period energy intensity of the ith sector by base period energy

intensity of the same sector, i.e., (IiT/Ii

0). The sum value of three

Equations (5a), (5b) and (5c) must be equal to the term ∆E= ET - E0 in

Equation (5).


4. RESULTS AND DISCUSSIONS

The study is carried out for three time periods using methodo-

logy described in Section 4. First analysis gives us overall energy

consumption decomposition for the whole period of 1990-2013. Table 1

shows the decomposition results for all three periods. In each period, the

first row show decomposition results in absolute form with unit of energy

consumption change, while second row shows the relative contribution

of the three effects in total change in percentage form. The total change

in energy consumption is 18.603 MTOE by taking 1990 as base year and

2013 as current year. The dominant contribution of 159.97% comes from

activity effect with 29.780 MTOE.

Intensity effect is the second dominant contributor having a share

of -12.722 MTOE, implying 68% reduction in overall change in energy

consumption, thus compensating the inflated demand of energy use

driven by activity and structural effects. It means that energy intensity

has decreased leading to a contraction of 12.722 MTOE in overall energy

change. Since energy intensity is inversely related to energy efficiency,

a decrease in energy intensity implies a rise in energy efficiency. Several

factors that contribute directly and indirectly towards energy efficiency

are innovative energy saving policies such as promotion of energy

efficient appliances, growth of natural gas consumption and LPG in all

the three sectors of economy during the period.

The third contribution of 8.41% with 1.565 MTOE in total

change comes from structural effect, showing that structure of the

economy experienced slight transition from agrarian to industrial and

services dominated economy. The sectoral composition of the economy

changed especially in the 2000s. The share of agriculture in economy

falls by 17% between 1990 and 2013. The shares of industry and services

sector increase by 10% and 5% respectively during the period. Statistics

show structural shift from low energy intensive agriculture sector to high

energy intensive industrial and service sector thus contributing in rise of

energy consumption.

The overall increase of energy consumption is dominantly

contributed by services and industrial sector. The cumulative effects of

service sector indicate that this sector is responsible for 12.501 MTOE


in total change. Similarly, industrial sector is responsible for an increase

of 6.937 MTOE in total energy consumption over the period 1990-2013.

Agriculture is responsible for a slight decline of 0.835 MTOE in the

overall change. The sector wise decomposition analysis shows a decline

of -0.835 MTOE in agriculture sector. Activity effect is responsible for

an increase of 1.022 MTOE due to growth in agriculture sector, while

structural and intensity effects dominantly decrease the total change.

Structural pattern also indicates within sector transition from energy

intensive crop sector towards less energy intensive livestock sector.

A rise of 6.937 MTOE in industry is mainly due to activity effect

followed by structural effect having share of 150% and 14% respec-

tively. Due to growth in industrial sector, activity effect is the main

source of growth in energy consumption. Structural effect shows that

industrial sector has been changing its sub-sector composition with rising

share of mining, quarrying and manufacturing. The compensating share

of 63% is reflected through intensity effect with decline of 3.689 MTOE

in total industrial energy consumption change. It means industrial sector

has witnessed energy efficiency due to modern techniques of production,

change in sub-sectoral composition with slight shift from more energy

intensive sub-sectors to less energy intensive sub-sectors as mentioned

in Figure 5.

An increase of 12.501 MTOE in services sector is mainly

attributed to activity and structural effects with share of 147% and 7%

respectively. Due to growth in services sector, activity effect is

responsible for 18.325 MTOE rise in total sectoral energy consumption

change. The structural changes in sub-sector composition of services

sector as shown in Figure 6, is captured by a rising share of sub-sectors

like finance and insurance and social and community services in total

service industry over the period 1990-2013. Intensity effect has

compensating share of 53% causing total energy consumption change

decrease by 6.56 MTOE. Energy efficiency in services sector is mainly

attributed to shift from high energy intensive subsectors to less energy

intensive subsectors. The dominant share of service sector in total energy

use implies that energy is used as a final good to maximize well-being

although it may not directly generate economic activity.


Table 1 shows the results of total change in energy consumption

which stands at 3.70 MTOE over the period 1990-2000. The dominant

factor behind this change is activity effect followed by intensity effect.

Activity effect (∆EAct) is responsible for 278% increase in total change

equivalent to 10.28 MTOE. The second dominant contributor is intensity

effect (∆EInt) with -6.51 MTOE holding a share of -175.85% in overall

change. The contribution of structural effect (∆EStr) is negligible with

0.066 MTOE (1.77%) decline in overall energy consumption change.

Figure 3 shows that sectoral composition of Pakistan economy in the

1990s was traditional with little structural shifts towards industriali-

zation. The negative contribution of structural effect in overall energy

consumption change is due to slight decrease in the service sector share.

To summarize, the total change of 3.703 MTOE is dominantly contri-

buted by services and industry. The cumulative effects of services sector

is 3.179 MTOE followed by industrial with 1.344 MTOE in total

increase. Agriculture sector is responsible for a decrease of 0.820 MTOE

in overall change in this period.

Table 1. LMDI Decomposition Results of Pakistan Economy:

1990-2013

Category (Unit) ∆EAct ∆EStr ∆EInt ∆ETOT

1990-2013

Absolute change

(MTOE) 29.78 1.565 -12.722 18.603

Share in change (%) 159.97 8.41 -68.38 100

1990-2000

Absolute change

(MTOE) 10.282 -0.066 -6.513 3.703

Share in change (%) 277.62 -1.77 -175.85 100

2001-2013

Absolute change

(MTOE) 17.98 1.900 -4.986 14.90

Share in change (%) 120.71 12.75 -33.46 100


The structural and intensity effect results for the Period 2000-

2013 are different as compared to preceding period. Structural effect has

a higher share of 12.75 % in the energy consumption change as compared

to -1.77 % in 1990s which clearly shows that economy had experienced

structural changes. The share of intensity effect falls from 176% in the

1990s to just 34% in the 2000s, which shows that energy efficiency still

provide relief to inflating energy consumption but its share is declining.

During 2000-2013, total energy consumption increases by an

amount of 14.90 MTOE. Due to fast pace economic growth in the mid-

2000s, the activity effect is the major contributor having share of 17.98

MTOE (121%). The share of activity effect in overall energy

consumption falls from 278% to 114% during 2000-2013 as compared

to 1990-2000 implying that rapid growth was absorbed and spread by

structural and intensity effects. The share of structural effect in total

change has increased from 1.77% in 1990s to 13% in this period.

Intensity effect is the most considerable factor in this period, showing

low trends of energy efficiency. In this period, intensity effect is

responsible for a decline in overall energy consumption change

equivalent to 4.98 MTOE (33%).

Table 2 shows sector-wise decomposition results. The agriculture

sector is responsible for 0.820 MTOE decline in total energy

consumption change mainly due to structural and intensity effects.

Activity effect explains the change in energy consumption by 0.45

MTOE (55%) mainly due to growth in agriculture sector. Intensity effect

has compensating share of 1.28 MTOE (157%) using decrease in total

agricultural energy consumption. Energy efficiency is mainly due to shift

from high energy intensive crop sector to low energy intensive livestock

sector along with the modernized farming techniques and energy

efficient farm machinery. The table shows a rise of 1.344 MTOE in

industrial sector’s energy consumption, due to activity and structural

effects and compensating intensity effect. The activity effect is

responsible for 262% increase in energy consumption accruing from

growth in the industrial sector. Structural effect accounts for 6% share in

total industrial energy consumption change. The industrial sector has

witnessed energy efficiency as depicted by a compensating share of

168% in total change. Modern production techniques, energy saving


machineries, high value added output and changing sub-sector com-

position of industrial sector are the main reasons of energy efficiency in

the industrial sector.

An increase of 3.179 MTOE in services sector is mainly due to

activity effect with a share of 198.43%. Structural and intensity effect

has compensating share of 5% and 93% respectively due to changing

sub-sectoral shares. Last panel of Table 2 shows an increase of 14.90

MTOE in 2000-2013, which is dominantly contributed by services and

industrial sector. The combined effect of services sector is 9.32 MTOE

followed by industrial sector with 5.59 MTOE over this period. The

effect of agriculture sector is negative with a decrease of 0.015 MTOE

in overall energy consumption change.

Table 2. Sector wise LMDI Decomposition Results of Pakistan

Economy 1990-2013

Sector ∆EAct

(MTOE)

∆EStr

(MTOE)

∆EInt

(MTOE)

∆ETOT

(MTOE)

∆EAct

%

∆EStr

%

∆EInt

%

Total

%

1990-2013

Agriculture 1.02 -0.209 -1.59 -0.84 -122.35 23.93 198.41 100

Industry 10.41 0.938 -3.68 6.94 150.09 13.52 -63.62 100

Services 18.32 0.827 -6.56 12.50 146.58 6.61 -53.20 100

Total (MTOE) 29.76 1.565 -11.84 18.60 _ _ _ _

1990-2000

Agriculture 0.45 0.013 -1.28 -0.82 -55.47 -1.65 157.12 100

Industry 3.51 0.081 -2.25 1.34 261.63 6.35 -167.99 100

Services 6.30 -0.162 -2.96 3.18 198.43 -5.17 -93.25 100

Total (MTOE) 10.28 -0.064 -6.51 3.70 _ _ _ _

2001-2013

Agriculture 0.37 -0.139 -0.25 -0.015 -2483 927.29 1655.75 100

Industry 6.28 0.891 -1.58 5.59 112.33 15.94 -28.28 100

Services 11.33 1.148 -3.16 9.32 121.53 12.31 -33.85 100

Total (MTOE) 17.98 1.900 -4.98 14.90 _ _ _ _


Sector-wise decomposition results show that compensating 0.015

MTOE change in the agriculture sector is attributed to structural effect

(927%) and intensity effect (1656%). Activity effect is responsible for

an increase of 0.37 MTOE in total change in agricultural sector energy

consumption due to the sectoral growth. Structural shifts are responsible

for decrease of 0.14 MTOE. Intensity effects also have compensating

share of 0.24 MTOE in the total change arising mainly from change in

sub-sector composition. The activity and structural effects with share of

112% and 16% respectively account for an increase of 5.59 MTOE in

the industrial sector. Activity effect is responsible for an increase of 6.28

MTOE in total industrial energy consumption change due to growth in

that sector. Intensity effect has compensating share of 3.16 MTOE,

indicating energy efficiency achievements in industrial sector. A rise of

9.32 MTOE in services sector is contributed by activity and structural

effects with share of 122% and 12%, respectively. Activity effect brings

about 11.33 MTOE increase in total service sector energy consumption

change due to massive growth particularly in the mid-2000s. Change in

sub-sector composition is responsible for an increase of 1.148 MTOE

through structural effect. Energy efficiency brings a compensating share

of 3.16 MTOE reduced through the intensity effect.

5. IMPACT OF ENERGY IMPORTS ON PAKISTAN

EXTERNAL ACCOUNT

Pakistan is a net energy importer and relies both on domestic and

imported resources to meet its requirements. Natural gas, oil and

hydroelectricity account for about 90% of energy supply in the country.

The government regulates the distribution of almost all major energy

sources. Natural gas supply is made out of domestic resource and often

its supply falls short of demand especially during the peak demand

season. In that case, oil or LNG imports satisfy excess demand and bring

the market into equilibrium. However, heavy share of imports in total

energy consumption puts pressure on the balance-of-payments account.

Some earlier studies highlight the impact of energy imports on the

external account of Pakistan [Malik, 2007; Jamil (2012)]. The depend-

ence on imported energy will further increases due to depleting domestic


resources in future that will create negative impact on country’s external

account. Jamil (2012) argues that internal energy policy and regulations

are causing a shortage of indigenous energy resources that resulted in

increased imports earlier than it is binding. Hussain (2010) also argues

that planning of energy supply composition should integrate the external

cost of energy use.

Sustainable and affordable energy supply drives to economic

growth and well-being. The rising demand for fossil fuels and its price

volatility put extra burden on energy importing economies and make

energy sector management and planning quite challenging. Energy

security especially in uncertain international oil market has been an

important political and economic concern. Energy dependence to fulfill

the total energy requirement can be measured as net energy import as

percent of total energy consumption. High energy dependence leads a

country to vulnerability with negative impacts on foreign exchange

reserves and expose the economy to international economic shocks.

The world is facing serious challenges to sustainably meet its

growing energy requirements and Pakistan, being an energy deficient

country, finds it hard to ensure reliable energy supplies. The rising

energy demand coupled with limited and depleting domestic energy

resources has unfolded since the mid-2000s. According to SBP Annual

Report 2012-13, Pakistan spent approximately USD 14.9 billion (33%

of total imports) on petroleum imports with a quantity of 17.9 MTOE.

According to Pakistan Integrated Energy Plan 2012-2025, if GDP grows

at the rate 2-4%, then the final energy consumption of Pakistan will reach

142 MTOE by 2025 with cumulative growth rate of almost 250%

(equivalent to 40 MTOE).

For many years, the country meets only 20% of its total oil

requirements through domestic production and the rest is imported.

Keeping the same ratio of 80:20 till 2025 implies that the country will

require to import 32 MTOE of oil which is equivalent to 220.8 million

barrels per annum. The forecasted world oil price for 2025 is $100/barrel,

taken from the World Bank commodity price index. At price $100/barrel,

Pakistan will have to spend $22.85 billion on oil imports by 2025. It is

pertinent to note that any increase in price of crude oil may further

increase the import bill.


According to Pakistan Integrated Energy Plan 2012-2025, by the

year 2025, Pakistan will also require 10.12 billion cubic feet per day

(BCFD) of natural gas with expected domestic production of 2.17

BCFD. The imports of 7.95 BCFD will be equivalent to 2992 million

British thermal Unit (MBTU) per annum. If the international price of gas

forecasted by the World Bank Commodities Price Forecast remains

$10/BTU, Pakistan will have to spend $ 29.92 billion by 2025. Table 3

shows the impact of projected imports on external account in 2025.

Projected estimates show that the country will have to spend $52.76

billion on imports of oil and gas by 2025. The total value of oil and gas

imports in 2013 were $14.9 billion that will witness a growth of 250%

by 2025.

Table 3. Import Bill based on Projections

Domestic Imports Total World Price Value

($ billion)

Oil 2025

(MTOE)

8

32 40

$ 100/barrel 22.85

Gas 2025

(BCFD)

2.17

7.95 10.12 $

10/(MBTU)

29.91

Oil 2013

(MTOE)

11.3 19 30.3 $ 107/barrel 14.9

Sources: Pakistan Integrated Energy Plan, 2012-2025, Pakistan Energy Year Book, HDIP, 2013, WB Commodity

Price Forecast, 2015.

High dependence on imported energy will further expose the

economy to vulnerability with interrupted energy supplies and price

shocks especially in international oil market. The negative impacts of

$52.76 billion worth imports will severely reflect through both internal

account (budget) and external account (balance of payment). The ever

rising-energy imports have serious implications for an economy with

trade deficits. In consideration of limited foreign exchange reserves,

there is a need to reduce our dependence on energy imports by domestic

resource mobilization, through inflow of foreign direct investment,

efficient use of available energy resources, and balanced energy mix.

Energy conservation is the best method to lower the energy supply and

demand deficit. Currently, Pakistan has the potential to save 25 % of

energy on average worth $5billion annually [ENERCON (2015)].


6. CONCLUSION

This study provides a comprehensive analysis of the factors

responsible for growing energy consumption in Pakistan over the period

1990-2013 with segmented time period analysis using LMDI

decomposition technique. Finding the determinants of change in overall

energy consumption is important for appropriate energy policy design

and its implementation. In Pakistan energy mix is mainly comprised of

four fuel types; natural gas, oil, hydroelectricity and coal. LMDI techni-

que tracks down the pre-defined determinants (activity effect, structural

effect and intensity effect) responsible for increased energy consumption

in three major sectors.

Total energy consumption in Pakistan increased by 86.2%

(equivalent to 18.60 MTOE) over the period 1990-2013. Even so, there

is prevalence of enormous unmet energy demand due to natural gas and

electricity shortage especially since the mid-2000s. The energy shortage

is a result of the domestic energy market control by the government, who

are responsible for overall energy planning and influence the prices and

quantity supplied to different sectors. This increase is mainly due to the

activity effect followed by an intensity effect. By 2025, the final energy

consumption is projected to reach 142 MTOE level. In analysis of the

period 1990-2013, activity effect is found to be the main contributor in

overall energy consumption growth, which reiterates the positive

relationship between economic growth and energy consumption.

Intensity effect is the second major compensating contributor in the

overall energy consumption change showing strong indications of

energy efficiencies in overall economy as well as in sector wise analysis.

The share of structural effects are low in volume but indicates that the

economy experiences a slow pace structural shift from traditional

economy to industrialized and service-led economy.

Analysis of the period 1990-2000 shows the highest share of

energy efficiency through intensity effect in the overall change. Activity

effect remains the most dominant contributor in the overall energy

consumption change with services and industrial sectors holding the

dominant share in overall activity effect. The share of structural effects

is very little due to stagnant sectoral composition of economy over this


period. The compensating share of intensity effect in total energy

consumption is comparatively less than the preceding period showing an

overall slowdown energy efficiency.

The quantification of energy imports based on projections shows

that Pakistan faces serious challenge of rising import bill due to

dwindling domestic energy reserves and energy price volatility. The

findings of our study are consistent with various past studies conducted

for different countries mentioned in literature review [Ediger and Huvaz

(2006); Mairet and Decellas (2009); Sahu and Narayanan (2010); Zhang,

et al. (2011); and Nasab, et al. (2012)]. The decomposition findings of

Pakistan economy is consistent with other developing countries in which

activity and intensity effect contribute dominantly with structural effects

up to some extent. The findings of this study provide basic foundations

for energy related policies both in their formulation as well as

implementation. For policy makers, the result of energy intensity effect

is startling as it shows an increase in energy intensity during recent

decade, which is a sign of concern. In order to reduce gap between

energy demand and supply, energy conservation is the best and cheapest

method as Pakistan has potential to save sufficient energy. Furthermore,

the petroleum exploration policies should be designed judiciously to

attract investment in order to reduce import dependence.

Annexure

List of Selected Studies on Energy Decomposition

Study Country Method Sector Effects/Factors

Choi, et al. (1995) Korea Divisia Index Manufacturing Aggregate

intensity

Ang and Lee (1996) Singapore

Taiwan

Energy Coefficient

Approach Electricity

Sectoral

intensity

Ediger and Huvaz (2006) Turkey LMDI Whole

Economy Sectoral

Ma and Stern (2008) China LMDI Whole

Economy Technological

Mairet and Decellas

(2009) France LMDI Services Activity

Sahu, and Narayanan

(2010) India

General Parametric

Divisia Index and

Laspeyers

Decomposition Index

Manufacturing

Structural

Intensity

Activity

Zhang, et al. (2011) China LMDI Transport Activity

Intensity


Balezentis and

Strienikiene (2011) Lithuania LMDI

Whole

Economy Intensity

Nasab, et al. (2012) Iran LMDI Industrial

Activity

Structural

Intensity

Hasabeigi and Sathaye

(2012) California LMDI Industrial Structural

Zhao, et al. (2012) China LMDI Residential

Income

Structural

Intensity

Activity

Population

Lotz and Pouris (2012) South

Africa LMDI

Whole

Economy Structural

Ullah, et al. (2014) Pakistan Fischer Ideal Index

(IDA)

Whole

Economy Intensity

REFERENCES

Ang, B. W. and F. Q. Zhang (2000) A Survey of Index Decomposition

Analysis in Energy and Environmental Studies. Energy, 25:12,

1149-1176.

Ang, B.W. (2004) Decomposition Analysis for Policy Making in

Energy: Which is the Preferred Method? Energy Policy, 32,

1131-1139.

Ang, B.W. (2005) The LMDI Approach to Decomposition Analysis: A

Practical Guide. Energy Policy, 33:7, 867-871.

Ang, B.W. and K. H. Choi (1997) Decomposition of Aggregate Energy

and Gas Emission Intensities for Industry: A Refined Divisia

Index Method. The Energy Journal, 18:3, 59-73.

Ang, B.W. and P.W. Lee (1996) Decomposition of Industrial Energy

Consumption: The Energy Coefficient Approach. Energy

Economics, 18, 129-143.

Ang, B.W. and F.L. Liu (2001) A New Energy Decomposition Method:

Perfect in Decomposition & Consistent in Aggregation. Energy,

26, 537-548.

Ang, B.W. and N. Liu (1995) Decomposition Methodology in Industrial

Energy Demand Analysis. Energy, 20:11, 1081-1095.

Ang, B.W. and N. Liu (2007) Negative-Value Problems of the Log Mean

Divisia Index Decomposition Approach. Energy Policy, 35:1,

739-742.


Ang, B.W. and F.Q. Zhang (2000) The LMDI Approach to

Decomposition Analysis: A Practical Guideline. Energy Policy,

33:7, 865-871.

Ang, B.W., Liu, F.L. and S.H. Chung (2004) A Generalized Fisher Index

Approach to Energy Decomposition Analysis. Energy

Economics, 26:5, 757-763.

Apergis, N. and J.E. Payne (2010) Energy Consumption and Growth in

South America: Evidence from a Panel Error Correction Model.

Energy Economics, 32:6, 1421-1426.

Balezentis, A., Balezentis, T. and D. Streimikiene (2011) The Energy

Intensity in Lithuania during 1995-2009: A LMDI Approach.

Energy Policy, 39:11, 7322-7334.

Belke, A., Dobnik, F. and C. Derger (2011) Energy Consumption and

Economic Growth: New Insights into the Cointegration

Relationship. Energy Economics, 33:5, 782-789.

Ediger, V.S. and O. Huvaz (2006) Examining the Sectoral Energy Use

in Turkish Economy: (1980–2000) With the Help of

Decomposition Analysis. Energy Conservation and

Management, 47:6, 732-745.

ENERCON (2015) Government of Pakistan. Available at: http://www.

enercon.gov.pk/index.php?option=com_content&view=article&

id=28&Itemid=27.

HDIP (2000, 2014), Energy Yearbook, Hydrocarbon Development

Institute of Pakistan, Government of Pakistan, Islamabad.

Hasanbeigi, A., Can, S.D.L. and J. Sathaye (2012) Analysis and

Decomposition of the Energy Intensity of California Industries.

Energy Policy, 46, 234-245.

Hatzigeorgiou, E., Polatidis. and D. Haralambopoulos (2008) CO2

Emissions in Greece for 1990-2002: A Decomposition Analyses

& Comparison of Results Using the Arithmetic Mean Divisia

Index & Logarithmic Mean Divisia Index Techniques. Energy,

33:3, 492-499.

Hoekstra, R. and J.J.C. Bergh (2003) Comparing Structural and Index

Decomposition Analysis. Energy Economics, 25:1, 39-64.


Hussain, T. (2010) Pakistan's Energy Sector Issues: Energy Efficiency

and Energy Environmental links. The Lahore Journal of

Economics, 15:3, 33-59.

Ministry of Finance (2013) Integrated Energy Planning and Pakistan's

Energy Future: An Update on a Continuing Crisis. Pakistan

Economic Forum, Islamabad.

Jamil, F. (2012) Impact of Different Public E&P Policies on Natural Gas

Reserves and Production in Pakistan. Resources Policy, 37:3,

368-374.

Jamil, F. and E. Ahmad (2010) The Relationship between Electricity

Consumption, Electricity Prices and GDP in Pakistan. Energy

Policy, 38:10, 6016-6025.

Lee, C. C. (2007) Energy Consumption and Economic Growth in Asian

Economies: A More Comprehensive Analysis using Pannel Data.

Resource and Energy Economics, 30:1, 50-65.

Liu, F.L. and B.W. Ang (2003) Eight Methods for Decomposing the

Aggregate Energy Intensity of Industry. Applied Energy, 76:1,

15-23.

Lotz, R.I. and A. Pouris (2012) Energy Efficiency in South Africa: A

Decomposition Exercise. Energy, 42:1, 113-120.

Ma, C. and D.I. Stern (2008) China’s Changing Energy Intensity Trend:

A Decomposition Analysis. Energy Economics, 30:3, 1037-

1053.

Mairet, N. and F. Decellas (2009) Determinants of Energy Demand in

the French Service Sector: A Decomposition Analysis. Energy

Policy, 37:7, 2734-2744.

Malik, A. (2007) How Pakistan is Coping with the Challenge of High

Oil Prices. The Pakistan Development Review, 46:4, 551-575.

Masih, A. M. and R. Masih (1996) Energy Consumption, Real Income

and Temporal Causality: Results From a Multi-Country Study

Based on Cointegration and Error-Correction Modelling

Techniques. Energy Economics, 18:3, 165-183.

Ministry of Finance (1991, 2001 & 2014), Pakistan Economic Survey,

Ministry of Finance, Government of Pakistan, Islamabad

Nasab, E.H., Aalami, R., Dahr, S.F. and M. A. Sadeghzadeh (2012) An

Analysis of Energy Consumption in Transportation & Industrial


Sectors- A Multiplicative LMDI Approach with Application to

Iran. Iranian Economic Review, 16:32, 1-17.

Reddy, B. S. and B. K. Ray (2010) Decomposition of Energy

Consumption and Energy Intensity in Indian Manufacturing

Industries. Energy for Sustainable Development, 14:1, 35-47.

Sahu, S. K. and K. Narayanan (2010) Decomposition of Industrial

Energy Consumption in Indian Manufacturing: The energy

intensity approach. Journal of Environmental Management &

Tourism, 1:1, 22.

Shahbaz, M., Zeeshan, M. and T. Afza (2012) Is Energy Consumption

Effective to Spur Economic Growth in Pakistan? New Evidence

from Bounds Test to Level Relationship & Granger Causality

Tests. Economic Modelling, 29:6, 2310-2319.

Soytas, U. and R. Sari (2003) Energy Consumption and GDP: Causality

Relationship in G-7 Countries and Emerging Markets. Energy

Economics, 25:1, 33-37.

Ullah, A., Khan, K., and M. Akhtar (2014) Energy Intensity: A

Decomposition Exercise for Pakistan. Pakistan Development

Review, 53:4, 531-548.

Xu, X.Y. and B.W. Ang (2014) Analyzing Residential Energy

Consumption using Index Decomposition Analysis. Applied

Energy, 113, 342-351.

Zhang, M., Li, H., Zhou, M. and H. Mu (2011) Decomposition Analysis

of Energy Consumption in Chinese Transportation Sector.

Applied Energy, 88:6, 2279-2285.

Zhao, X., Li, N. and C. Ma (2012) Residential Energy Consumption in

Urban China:A Decomposition Analysis. Energy Policy, 41,

644-653.



The Productivity Paradox: Does Happiness Matter?

Farooq Rasheed* and Eatzaz Ahmad†

Abstract:

Productivity paradox refers to a situation when investment in information

technology is inversely related with economic growth. We test the hypothesis whether

economic happiness plays a moderating role between information related

capitalization and economic growth and thus may or may not invalidate the

productivity paradox. Using 20 years annual panel data of OECD and APT economic

blocs, happiness moderates the relationship between inputs labour and capital

productivities with output growth in both the economic blocs. Economic Happiness is

thus recommended to be boosted in getting effective labour and capital productivities.

Keywords: Productivity Paradox, Happiness, Information Technology

1. INTRODUCTION

a. Productivity Paradox

Economic inefficiencies are often seen in various regions of the

world. It is observed that economic growth often lags the amazing

technological progress and South Asian region is no exception. Islam,

Salim and Bloch (2016) examine the impact of intra-regional initiatives

on various aspects of efficiency along with growth of productivity in

South Asia. They observed that the South Asian region has suffered from

a total factor productivity shrinkage and economic hammering of some

degree through technological inefficiencies and are slow in adopting

such technological innovations.

Amjad and Awais (2016) review Pakistan’s productivity

performance over the period of 1980 to 2015. Authors examined the

contribution of physical capital, human capital and TFP to labour

* Mr Farooq Rasheed <[email protected]> is Assistant Professor at Air

University, Islamabad, Pakistan. † Eatzaz Ahmad <[email protected]> is Chair Professor of the State Bank of

Pakistan at University of Peshawar, Peshawar, Pakistan.


179 Rasheed and Ahmad

productivity and observed that the contributions of physical capital and

education remained modest and there has been a declining trend in TFP

growth. They identified the declining trend in labour productivity and

total factor productivity (TFP) for which the lack of sustained growth

and declining levels of technological investment were found to be the

key causes. They concluded that Pakistan’s economy has not taken full

advantage of the favourable technological developments. This is

somewhat similar to the predictions of Moore’s Law, which has held for

more than four decades. It was noted that companies bought computers

on the guarantee that the “computer age” would revolutionize business.

In the 1970s, information technology (IT) related technical

equipment accounted for about 25% of all information technology

related business investments. However, a number of researches in the

1980s and 1990s failed to get any evidence for improvement in such a

technological productivity contribution [Berndt and Morrison (1995)].

In the 1980s and the early 1990s, the “productivity paradox” was widely

debated. The productivity paradox is the unusual observation made that

as more investment is made in IT, workers’ productivity instead of

growing actually declines. Despite striking advances in computer field

and increasing capacity of the IT sector, growth rate of productivity

declined in the US economy. Labour productivity growth rate in the

1960s was around 3% and fell to approximately 1% in the beginning of

the new millennium. These paradoxical productivity patterns are also

referred to or termed as “Solow Computer Paradox” due to Solow’s

(1987) statement “You can see the computer age everywhere but in the

productivity statistics”. Researches point out three possibilities for such

a paradox:

i. Data and analytical problems hide productivity revenues, i.e., the

ratios for input and output are not easy to measure. As a culture

moves progressively from a qualitative one to a quantitative one, its

effect on productivity increases even more, thus further hiding the

gains which could be ascribed to technology.

ii. Revenues gained by a company through productivity may not always

be easy to account for, because these could be offset by losses in

other divisions or departments. But as the overall productivity is

https://en.wikipedia.org/wiki/Computer_age

The Productivity Paradox: Does Happiness Matter? 180

considered, these could be buried in the details. Once again, profits

accrued just by investments in productivity become hard to measure.

iii. Complex designing, administering and maintaining of IT systems.

These costs are due to rapid obsolescence of equipment and software,

incompatible software and network platforms and issues with

security such as data theft and viruses. This drives a continuous cycle

of technology replacement.

Wetherbe, et al. (2007) emphasize that in order to interpret the

paradox, the concept of productivity has to be well understood. The

existence of the paradox may not be the same among different firms or

economies. This could be due to the differences in their efforts of

adopting technological development and possibly the differences in

standards of information technology.

To answer how information technology affects productivity, it

seems important to understand the functional role of information

technology towards production. Economists consider its role as a factor

of production and thus provide vital statistics for growth accounting and

help in defining meaningful linkage between inputs like labour, capital

and IT capitalization with growth rate of the output of an economy. Table

1.1 showing the trends and reoccurrence of such a paradox is visible

beyond year 2001.

Most of the studies have also found that in comparison to other

capital investment, IT investments were associated with higher marginal

productivity. Some studies translate these returns into “excess returns”,

by stressing a perspective that investments should pay the same risk

adjusted returns thus ending up with lower net returns. In contrast,

Brynjolfsson and Hitt (1996) observed that the net returns from IT

investments were nonetheless more than the returns in non-IT

investments; partly due to the complementary nature of such investments

that the firms establish for raising assets.

Acemoglu, et al. (2014) have revisited the productivity growth

and IT related issue and found no significant evidence of increasing

productivity growth in labour working in IT exhaustive industries-

bearing high cost of technological investments. Furthermore, authors are

of the view that such IT related investments are one of the causes of


workplace distractions. Arguably this could have negative effects at

workplace productivity till such noise pollution factors are controlled.

Table 1.1. Reoccurrence of Productivity Paradox

In their study on information technology business value,

Melville, et al. (2004) split performance into two categories: business

process performance and organizational productivity and define that

business process related performance is actually operational efficiency

measured by factors like customer service, flexibility and culture of

information sharing, while the factors like market value, profitability,

competitive advantage account for firm’s productivity. Sunny, et al.

(2005) noted significant association between IT investments and

performance efficiencies in the hotel industry through factors like

enhanced annual sales, greater level of repeat business, increased

occupancy rate, enhanced positive word of mouth and reduced operating

costs. Gartner (2012) states “However, despite unclear causality, on an

industry level there appear to be interesting relationships between the

level of investment and the operating profits of organizations… many

organizations with high operating margins also have high IT spending as

a percentage of revenue. This view should not imply that, by investing

more in IT, an organization should expect to get better profitability,


rather, it should simply outline how different industries behave under

varied economic conditions.” This suggests that financial return cannot

be considered as the sole indicator of performance, rather other factors

are required to be explored.

b. Economic Happiness

Kazi (2002) using Figures 1 and 2 demonstrates how the

technological reliance and economic dependence are complementary in

nature, but often forms a vicious circle of technological reliance. This

vicious circle encircles increased dependence on foreign inputs that lead

to weak indigenous capacity and thus an ineffective economic growth

and development. It may result in possibilities like retreating economic

growth, enhancing poverty index, reduction in wellbeing index for the

society, etc. The idea that wellbeing or happiness is becoming important.

Magnus (2001) using Swedish micro data reported that the economics of

happiness is quantitative and theoretical in nature that studies well-being,

life satisfaction, quality of life, psychological and health aspects.

Though the wealth accumulation is considered as the key

objective in life, but should not be at the cost of happiness. Mostly the

measures of economic improvements do not reflect the happiness

aspects. Can money buy happiness? The debate on this query is now

taking place frequently. Anielski (2009) classifies assets as natural,

financial, human & social types and stresses to improve them to build

not just a wealth but a genuine wealth.

This helps recent literature also emphasize on happiness and

labor productivity linkage. The dimension of happiness has been taken

into account to test such linkage to support the happy society view of

philosophers like Aristotle, Confucius and Plato.

Does a rise in happiness affect productivity? Through conducting

three different styles of experiments, Oswald, Proto & Sgroi (2015)

found that happiness makes labors more efficient and productive. In the

experiments, the selected persons were made happier and found that the

treated people were 12% more productive and observed that lower

happiness is associated with low productivity.


Figure 1. The Vicious Circle of Technological Reliance

L o w D e m an d f o r I n d i g e n o u s S & T Se r v i c e s

W e ak I n d i g e n o u sC ap ac i t y

S & T

R e l i an c e o nF o r e i g n I n p u t s

M o r e F o r e i g nI n p u t s

I n e f f e c t i v e d e v e l o p m e n t o f I n d i g e n o u s S & T

F u r t h e r m ar g i n a lI n d i g e n o u s S & T

i sat i o no f

I n c r e ase d d e p e n d e n c eo n f o r e i g n I n p u t s

Source: Adopted from Kazi (2002)

To see the impact on economic growth, two variables - life

expectancy and investment ratio representing happiness were

considered to test the significance in the relationship between them, Li

& Lu (2010) found a robust positive correlation.

Guriev and Zhuravskaya (2009) in contrast found a positive and

significant effect on life satisfaction with Gini coefficient. Rasheed,

Ahmad and Rauf (2011) using pooled data of industrial / developed

economies, observed a significant impact of economic happiness on

GDP growth.

Veenhoven (2000) established counter logical connection

between economic happiness and income equality and found that the

presumed link fails to exist. Average economic happiness was high in

countries where income related equality was poor.


Figure 2. The Vicious Circle of Economic Dependence Superimposed

on the Circle of Technological Development

R e t ar t e d D e v e l o p m e n to f L o c a l I n d u st r y

H i g h E c o n o m i cD e p e n d e n c e

H i g h D e m an d f o rF o r e i g n L o an s

W e ak I n d i g e n o u s C ap ac i t y

S & T

M o r e I m p o r t s

L e ss R e so u r c e s f o rI n d i g e n o u s S & T

D e v e l o p m e n t

I n e f f e c t i v eD e v e l o p m e n t o f

I n d i g e n o u s S & TD e v e l o p m e n t

I n c r e ase d D r ag o nF o r e i g n E x c h an g e R e se r v e s

Source: Adopted from Kazi (2002)

Since a happy state of labour is more productive therefore a

happiness factor can influence the existing relationship between

capitalization in information technology (IT) and economic growth

through effective labour productivity and thus invalidates productivity

paradox. Happiness aspect as a policy objective in achieving higher

productivities is generally overlooked. This study thus attempts to test

the hypothesis whether economic happiness plays moderating role

between information related capitalization and economic growth.

2. DATA AND METHODOLOGY

In order to test the argument we raised, the annual time series

data of total employment to represent participatory labour force (N)1,

1 Expectedly, labour force in general is more the more IT literate over the time.


Investment in Information and Communication Technologies (ICT) to

represent capital and GDP(Y) at factor cost for the all OECD and

ASEAN plus three (APT2) countries are taken from The World Bank’s

data bank for a period of 20 years from 1996 to 2015. The happiness (H)

data are taken from world data base of happiness and New Economic

Foundation, UK. Generally in OECD countries the growth in investing

in information technology is higher and modern than in Asian

economies, therefore we have considered OECD and APT blocs to be

compared for the hypotheses we have developed.

The data are transformed into natural log form to get

responsiveness of each input, i.e., labour and capital towards output

productivity using panel least square regression estimates of the

following equations with happiness as a moderator in which Equations

2.1 and 2.2 will test the moderating role3 of happiness factor. Since

pooled ordinary least square (PLS) mitigates the existence of non-

independent observations, thus catering to the issue of serial correlation;

thus we preferred PLS over other regression models such as fixed or

random effect models.

Ln Yt = β0 + β1 Ln Kt + β2 Ln Nt+ εt …(2.1)

TFPt = λ0 + λ1Ln Ht +εt …(2.2)

Ln Yt = b0 + b1 Ln Kt + b2 Ln Ht + b3 Ln KtHt + εt …(2.3)

Ln Yt = d0 + d1 Ln Nt + d2 Ln Ht + d3 Ln NtHt+ εt …(2.4)

where,

Ln Yt = Natural Log of GDP at factor cost,

Ln Kt = Natural Log of Investment in Information and Communication

Technologies,

Ln Nt = Natural Log of Labour,

Ln Ht = Happiness index,

Ln HtKt and Ln NtKt are the cross product terms to test moderation,

2Consists of countries from ASEAN bloc plus China, Japan, and South Korea. 3Testing Moderation: Moderator effects are indicated by the interaction of an independent

variable and the moderating variable in explaining dependent variable. Following equation

demonstrates estimation of moderated regression Y = Z + aX + bM+ cXM + E. The interaction

of X and M measures the moderation effect via slope ‘c’. Slope ‘a’ measures the main effect of

X. The effect of X on Y is a + cM, thus, the impact of X on Y depends on M is also justified

[Baron & Kenny (1986)].


TFPt = Total factor productivity represented by the residual series ‘ε’ is

the error tem, Subscript‘t’ is the time period.

3. RESULTS

The panel regression estimates of the testing hypotheses are

provided in Appendix I. Since each country specific cross section

selected in the pooled data has different types of stochastic components

which means some countries may have a unit root and some may not. In

this situation we have applied common unit root test [Levin, Lin and Chu

(2002)] to test the stationarity of all the selected series (see Table 3a in

the Appendix I). The results show that non-stationarity in the series

found at levels and stationarity at the first difference of the selected

series, suggesting a possibility of long run policy implications.

Table 3.1. Results for Economic Growth

(OECD Bloc)

Variable Coefficient Prob. R-squared Durbin-Watson

C 11.15 0.00 0.96 1.12

Ln(K) 0.31 0.00

Ln(N) 0.43 0.00

(APT Bloc)

C 23.14 0.00 0.99 0.81

Ln(K) 0.53 0.00

Ln(N) 0.27 0.04

Note: Estimates are based on Equation 2.1.

For both OECD and APT countries, with strong degree of

coefficient of determination and no issue of autocorrelation, the response

of investment in information communication technology (ICT) on GDP

is found significant at 5% level of significance and the response of labour

on GDP is also found significant at 5% level of significance (see, Table

3.1).

TFP incorporates technological change, i.e., adopting new

technologies and happiness through sociopolitical factors [Bosworth and

Collins (2008)]. To test economic happiness factor as a determinant of

total factor productivity (TFP), we estimated Equation 2.2 using the


residuals from Equation 2.1 and found economic happiness as

statistically significant in both OECD and APT economic blocs at 10%

and thus gives a reason to test economic happiness factor as a moderating

variable (see, Table 3.2).

Table 3.2. Results for Total Factor Productivity

(OECD Bloc)

Variable Coefficient Prob. R-squared Durbin-Watson

C 19.30 0.34 0.99 2.88

Ln(H) 1.56 0.09

(APT Bloc)

C 1.25 0.12 0.98 2.18

Ln(H) 0.85 0.06


The possibility of moderation of economic happiness on

investment in ICT and GDP relationship is also proved significant at 5%

level of significance for OECD bloc and significant for APT bloc at 10%

level of significance (see, Table 3.3). Finally in the case of labour and

GDP association, the positive moderating role of happiness is again

statistically established for both OECD and APT blocs at 1% level of

significance (see, Table 3.4).

Table 3.3. Results for Moderator on IT

(OECD Bloc)

Variable Coefficient Prob.

C 5.08 0.00

Ln(K) 0.10 0.09

Ln(H) 0.29 0.00

Ln(KH) 0.16 0.03

(APT Bloc)


C 0.85 0.00

Ln(K) 0.02 0.00

Ln(H) 0.19 0.00

Ln(KH) 0.08 0.00



Table 3.4. Results for Moderator on Labour

(OECD Bloc)


C 3.31 0.00

Ln(N) 0.80 0.00

Ln(H) 1.12 0.00

Ln(NH) 0.19 0.00

(APT Bloc)


C 1.48 0.00

Ln(N) 1.21 0.00

Ln(H) 1.19 0.00

Ln(NH) 0.54 0.00

Note: Results are based on Equation 2.4.

4. CONCLUSION

Acemoglu, et al. (2014) observing growth patterns among IT

based manufacturing industries in US economy from 1980-2009, found

some evidence of productivity growth in IT based industries but not

beyond the 1990s; accompanied by observation of a steep rise in

unemployment. However, authors are of the view that rejection of Solow

Paradox by that proponents of the technological discontinuity view may

have been premature. We believe there can be some other factors that

can reverse the situation. We, by introducing economic happiness as an

intangible capital, claim that economic happiness can be augmented with

labour force to make them as effective labour force that can help in

raising production of efficiencies IT based. Our results approve that

economic happiness is pivotal in enhancing the impact of labour and

capital productivities on output growth.

The policy makers for an economy must aim at enhancing this

intangible capital which in the end heaves economic values. The vast

literature suggests quality of life, life satisfaction, social security,

religious independence, access to justice, equality, health, education are

vital determinants of economic happiness and among others thus may be

aimed at developing economic policies.


APPENDIX I

Table 3a Common Unit Root Process for all series

Method Statistic Prob.**

Unit root test at level

Levin, Lin & Chu t* 14.4001 1.0000

Unit root test at first difference

Levin, Lin & Chu t* -4.11307 0.0000

REFERENCES

Acemoglu, D., Dorn, D., Hanson, G. H., and B. Price, (2014) Return of

the Solow paradox? IT, Productivity, and Employment in US

Manufacturing. The American Economic Review, 104:5, 394-

399.

Amjad, R., and N. Awais (2016) Pakistan's Productivity Performance

and TFP Trends, 1980-2015: Cause for Real Concern. The

Lahore Journal of Economics, 21, 33.

Anderson, P. (1997) Review: Organizational Linkages: Understanding

the Productivity Paradox. Administrative Science Quarterly,

42:3, 595-597.

Anielski, M. (2009) The Economics of Happiness: Building Genuine

Wealth. Gabriola Island, British Columbia, Canada: New Society

Publishers.

Baron, R. M., and D. A. Kenny (1986) The Moderator–Mediator

Variable Distinction in Social Psychological Research:

Conceptual, Strategic, and Statistical Considerations. Journal of

Personality and Social Psychology, 51:6, 1173.

Berndt, E. R. and C. J. Morrison (1995) High Tech Capital Formation

and Economic Performance. Journal of Econometrics, 65, 9-43.

Bosworth, B., and S. M. Collins (2008) Accounting for Growth:

Comparing China and India. Journal of Economic Perspectives,

22:1, 45–66.

Brynjolfsson, E., and A. Saunders (2010) IT’s Contributions to

Productivity and Economic Growth. In Wired for Innovation:


How Information Technology Is Reshaping the Economy, 41-60,

MIT Press.

Gartner (2012) IT Metrics: IT Cost Transparency Action Plan, Definition

and Performance Metrics to Start the Journey.

Gil-Padilla, A. M. and T. F. Espino-Rodriguez (2008) Strategic Value

and Resources and Capabilities of the Information Systems Area

and Their Impact on Organizational Performance in the Hotel

Sector. Tourism Review, 63:3, 21–47.

Guriev, S., and E. Zhuravskaya (2009) (UN) Happiness in Transition.

Journal of Economic Perspectives, 23:2, 143-68.

Guzzo, R. A. (1996) Review: Organizational Linkages: Understanding

the Productivity Paradox. Academy of Management Review,

21:4, 1235-1237.

Islam, A., R. Salim, and H. Bloch (2016) Does regional integration affect

efficiency and productivity growth? Empirical evidence from

South Asia. Review Urban & Regional Development, 28: 107–

122. doi:10.1111/rurd.12048

Ives, B. (1994) Editor's Comments: Probing the Productivity Paradox.

MIS Quarterly, 18: 2, 21-24.

Kazi, M.A. (2002) “The Need for Integrated S&T policies for Islamic

Countries: The Role of Comstech” in Essays on Science. Edited

by Mohammad Said, Baital-Hikmah

Johnston, R. and P. Jones (2004) Service productivity: Towards

Understanding the Relationship between Operational and

Customer Productivity. International Journal of Productivity

and Performance Management, 53: 3, 201–213.

Lee, G. and J. Perry (2002) Are Computers Boosting Productivity? A

Test of the Paradox in State Governments. Journal of Public

Administration Research and Theory: J-PART, 12: 1, 77-102.

Li B. and Y. Lu. (2010) Accessed from

http://profile.nus.edu.sg/fass/ecsluyi/happiness_sub.pdf

Lichtenberg, F.R. (1995) The Output Contributions of Computer

Equipment and Personnel: A Firm Level Analysis. Economic

Innovations and New Technology, 3, 201-217.

Levin, A., Lin, C.F., and C.S.J. Chu (2002) Unit Root Tests in Panel

Data: Asymptotic and Finite-Sample Properties. Journal of

Econometrics, 108: 1, 1-24.


Lopez-Calix, J. R., Srinivasan, T.G., and M. Waheed (2012) What Do

We Know about Growth Patterns in Pakistan? (Policy Paper

Series on Pakistan No. 05/12) Washington, D.C.: World Bank.

Marchand, D.A. (2005) In Business Management. Improving

Performance using Information System. By Dave Chaffey &

Steeve Wood. Essex, UK PEL.

Melville, N., Kraemer, K. and V. Gurbaxani (2004) Review: Information

Technology and Organizational Performance: An Integrative

Model of IT Business Value. MIS Quarterly, 28:2, 283-322.

Moore, G.E. (1965) Cramming more components onto integrated

circuits. Electronics, 38: 8. Accessed on March 1st, 2017:

https://drive.google.com/file/d/0By83v5TWkGjvQkpBcXJKT1I

1TTA/view

Naym, J. and M.A. Hossain (2016) Does Investment in Information and

Communication Technology Lead to Higher Economic Growth:

Evidence from Bangladesh. International Journal of Business

and Management, 11: 6, 302.

Oswald, A.J., Proto, E., and D. Sgroi (2015) Happiness and productivity.

Journal of Labor Economics, 33: 4, 789-822.

Pinsonneault, A. and S. Rivard (1998) Information Technology and the

Nature of Managerial Work: From the Productivity Paradox to

the Icarus Paradox. MIS Quarterly, 22: 3, 287-311.

Preacher, K. J., Rucker, D. D., and A. F. Hayes (2007) Addressing

Moderated Mediation Hypotheses: Theory, Methods, and

Prescriptions. Multivariate Behavioral Research, 42: 1, 185-227.

Rasheed, F., Ahmad, E. and S.A. Rauf (2011) Happiness in Solow

Growth Model. Journal of Business and Economics, 3: 2, 127-

144.

Solow, R. (1987) Book Review: We’d Better Watch Out. New York

Times, July 12, 36.

Sunny, H., Kim, W.G. and S. Jeong (2005) Effect of Information

Technology on Performance in Upscale Hotels. International

Journal of Hospitality Management, 24: 2, 281-294.

Triplett, J. (1999) The Solow Productivity Paradox: What do Computers

do to Productivity? The Canadian Journal of Economics, 32: 2,

309-334.

https://drive.google.com/file/d/0By83v5TWkGjvQkpBcXJKT1I1TTA/view

https://drive.google.com/file/d/0By83v5TWkGjvQkpBcXJKT1I1TTA/view


Veenhoven, R. (2000) The Four Qualities of Life: Ordering Concepts

and Measures of the Good Life. Journal of Happiness Studies, 1:

1, 1-39.

Wetherbe, J. C., Turban, E., Leidner, D. and E. R. McLean (2007)

Information Technology for Management: Transforming

Organizations in the Digital Economy (6th ed.). New York:

Wiley.



Rehabilitation of 2010 Flood Affectees in Pakistan:

The Role of Development Partners

Sheeba Farooq*

Abstract:

Pakistan was hit by catastrophic floods in 2010, which inundated

infrastructure spread over 100,000 sq kms and affected over 20 million people with

large scale displacement. The study explored the impact of floods on the lives of people;

especially their livelihoods and gauged the extent to which the donors contributed

towards restoration of livelihoods of the flood affected population. This was

substantiated through a case study undertaken in Nowshera District, Pakistan, which

was amongst the worst hit areas during 2010 floods. The study used qualitative

research tools to make an in-depth analysis by gathering facts from the affectees and

validating it through the published reports. Semi-structured interviews were held with

the government functionaries, affected population and donor officials. The findings

indicate that almost seventy percent of the population received the Watan Card, which

helped them in resettling and rehabilitating their livelihoods but the rest thirty percent

were unhappy as they thought that the amount was not sufficient to re-activate their

livelihoods.

Keywords: Livelihoods, Rehabilitation, Watan Card

1. INTRODUCTION

1.1. Background

Located in South Asia, Pakistan is the sixth most populated

country in the world, housing a population of 184.35 million [GOP

(2013)] and stretching over an area of 700,000 sq km. Nature has blessed

the country with all endowments in abundance. Its vast expanse includes

long coastal areas, mineral rich deserts, fertile farmlands and lofty

mountains. Five of the fourteen tallest peaks including the second tallest

K-2 are located in Pakistan [Khan (2013)]. However, ever since its birth

* Sheeba Farooq <[email protected]> is Lecturer at School of Social

Sciences and Humanities (S3H), National University of Sciences and Technology

(NUST), Islamabad, Pakistan.

195 Sheeba Farooq

in 1947 the country has remained embroiled in territorial conflict with

neighboring India. In addition, Pakistan has also been facing continuous

ethnic and racial issues as well as political upheavals internally.

Resultantly, except for a few brief periods, the country has been ruled by

the military dictators. The Soviet occupation of Afghanistan, their

retreat, subsequent Talibanization and the post 9/11 war on terror has

taken a major toll upon the already meager economy of Pakistan and has

wrecked its socio-political fabric.

Pakistan has been enduring troubles throughout which has not

allowed its decision-makers to focus upon the issues of socio-economic

development. The lofty mountains, whose melting glaciers could be

utilized for hydel power generation and preservation of water resources

with some planning, now instead bring destruction to infrastructure and

losses to human lives by causing floods with increasing frequency.

1.2. Magnitude of Floods in 2010

Pakistan since its inception has experienced twenty floods1;

however the 2010 floods were the worst in its history. The United

Nations Secretary-General Ban Ki-moon highlighted the devastation

caused by 2010 flood by stating that, “in the past I have witnessed many

natural disasters around the world, but nothing like this” [Brooker

(2011:5)]. This flood according to him was a ‘slow-motion tsunami’. The

studies indicate that “heavy rainfall, flash floods and riverine floods

combined to create a moving body of water equal in dimension to the

land mass of the United Kingdom” [UNOCHA, (2010:10); World Bank,

(2011:7)].

The brutal impact of floods destroyed the lives of millions of

people living in 78 out of 141 districts in Pakistan [Brooker (2011)].

Reports produced by House (2012) and FAO (2012) illustrated that 20

million people and 1.74 million houses were affected by the floods. Not

only that it also brought large scale destruction to infrastructure, water

1As reported by the Ministry of Water & Power (2015), the floods from 1950 till

2015 have affected total area of 616,598 sq km, snatched 12,177 human lives and

caused direct losses worth US$38,165 million.

Rehabilitation of 2010 Flood Affectees 196

channels, sanitation, healthcare, housing and educational institutions

resulting in total disruption of life [UNOCHA (2010)].

The floods jolted the whole economy [Tahir, et al. (2011)].

People became homeless, lost their properties and instruments of income

generation and poverty level rose significantly. For instance, in Punjab

it reached 32.7% after 2010 floods as against 19.7% before floods and in

Khyber Pakhtunkhuwa it rose to 33.2% as against 19.4% pre-floods

[World Bank (2011)].

1.3. Destruction Caused by 2010 Floods in Pakistan

Government of Pakistan requested ADB and WB to carry out the

Damage and Needs Assessment (DNA) in the aftermath of the floods.

The findings of the report were that “the overall recovery and

reconstruction cost associated with the floods is estimated at approxi-

mately US$ 8.74 billion to 10.85 billion, which includes estimated costs

for relief, early recovery and medium to long-term reconstruction”

[World Bank (2010:24)].

The floods caused total or partial damage to almost 1.6 million

housing units and a total of 10,407 educational institutions in the country

were destroyed. Physical infrastructure was also marred by the floods as

estimates have indicated 10% of the road network and 16% railways

were damaged. The agriculture, livestock and fisheries sectors suffered

the highest damages estimated at US$5 billion [World Bank (2010)].

Hence, the floods caused destruction in all the sectors on a wide scale.

The study further highlights that the World Bank in the wake of

the floods initially proposed three options to reconstruct the damages

incurred by the floods as they were not in favour of giving cash assist-

ance. The first option was the Base case with a cost estimated at US$6.8

billion, second option was Building Back Smarter that would ensure

cost-optimization in reconstruction with a cost of US$7.4 billion and the

third was Building Back Better which comprised of larger infrastructure

improvements and was the recommended option by the World Bank with

a cost calculated at US$8.9 billion [World Bank (2010)]. But for one

reason or another and perhaps because of some political pressure the

government opted for cash assistance from donors rather than any of the

options mentioned above.

197 Sheeba Farooq

Government of Pakistan took immediate steps to provide

emergency relief which entailed distribution of essential items. But with

its meager resources and weak economy Pakistan was hardly prepared to

cope with the crises. It entailed rescue of the disaster-effected

population, provision of food and shelter, rehabilitating the people back

to pre-flood level and re-building the inundated infrastructure. Thus, the

country needed support in cash and kind from international donors which

did come.

How adequate and effective was the assistance extended by the

international community, is what this study intends to assess and

ascertain. The study estimates the wider debate of aid effectiveness in

the context of rehabilitation of disaster-effected population with

particular reference to the role of donors and their contribution during

2010 floods in Pakistan. The study shall also ascertain whether

international assistance has helped the Government of Pakistan in

preparing a disaster risk reduction strategy for future. Since the intensity

of floods in Pakistan seems to have rather increased, perhaps due to the

effects of climatic changes, it needs to adopt a pro-active approach

ensuring that the impact of natural disasters like floods is minimized and

that people whose socio-economic activities are disrupted are supported

to rehabilitate their livelihoods as fast and as effectively as possible.

Moreover, the issue of elite capture and its prevalence during the 2010

floods in various areas of Pakistan will also be highlighted in the study.

2. REVIEW OF LITERATURE

The territory of Pakistan comprises valleys and delta wherefrom

the mighty Indus River and all its tributaries flow. Floods in these rivers

are caused by heavy rainfall in upper catchments and Himalayan

foothills resulting from monsoon currents originating from Bay of

Bengal” [Tahir, et al. (2011:4)]. Sixty percent of Pakistan’s population

is still living below poverty line estimated according to the international

poverty line of two dollar a day [WDR (2013)]. The country is facing

multiple socio-economic problems and since 2008-09, the economy on

average grew at 2.94%. During FY12 and FY13 the power shortage

became so severe that it wiped out 2% of GDP [GOP (2013)]. Though


economy has the growth trajectory of more than 6%, but the worst energy

crisis, bleeding public sector enterprises, economic mismanagement and

menace of informal economy has hemorrhaged the system.

Some governments in the developing countries, with the help of

western donors and development partners are trying to build better

infrastructure to minimize the impact of such disasters but more often

than not the approach is that of a reactive nature rather than proactive

preparation. It is when the disasters have already played havoc with

people and infrastructures the governments wake up to the situation and

look for remedies with the meager resources available to the developing

world. It has been frequently observed that most of the developing

countries tend to make investments on recovery from a disaster rather

than creating adaptive capacity [Mirza (2003)].

The 2010 floods in Pakistan were according to all assessments,

of much greater magnitude than the 1998 floods in Bangladesh. As

earlier stated in World Bank (2011:7). The Government of Pakistan, civil

society and development partners such as the multilateral and bilateral

aid donors in the country, came forward with the best of their capacity

and efforts to save lives, provide immediate relief and support

rehabilitation of the population hit and displaced by the floods.

Since the major focus of this paper is to ascertain the role of

donors towards the rehabilitation of the flood affectees, we will start our

Literature Review by first discussing the wider debates such as Aid

Effectiveness and Disaster Risk Reduction and Livelihood Restoration

measures adopted in the developing countries, by applying a sustainable

livelihood framework to increase the resilience of the poor people. The

review will then ascertain the role of donors towards the rehabilitation

of the flood affectees. The review will study in depth the Citizen’s

Damage Compensation Program (Watan Card), the cash assistance

scheme launched by the Government of Pakistan with the support of the

development partners, and will assess its effectiveness towards the

rehabilitation of the flood affectees and restoration of their livelihoods.

During this discussion we will also throw light upon such themes as Elite

Capture, which prevails in most of the cash assistance scenarios.

A lot of aid has flowed to underdeveloped world after the World

War II and it perhaps has led to considerable socio-economic progress.

199 Sheeba Farooq

There is a continuing debate however, that has this aid been as effective

as it should have been in terms of value for each dollar. The governments

and the development agencies in the developed world are suffering from

aid fatigue and their tax payers are increasingly getting critical of their

tax money being thrown into ‘bottomless pits’ [OECD (2014)]. Have the

billions of aid dollars spent in Africa, South America and Asia been

effective to combat disease, illiteracy and poverty is the question being

asked at a number of foras with increasing frequency.

Masud and Yontcheva (2005) state that aid flows are predomi-

nantly meant to fill the gap between domestic savings and investment

needs. But drawing on reviews of the wider literature it has been argued

that aid has no significant impact on growth, savings or investment.

According to Mosley et al. (1992) aid indicated an increase in

unproductive public consumption and failed to promote investment but

then the proponents of aid have argued that developing countries would

have been poorer in the absence of aid. Gilbert and Vines (2006)

highlight the wider debates around aid which are conditionality and

dependency; the former has been criticized as the donor agencies; desire

to bring in policy reforms in poor environments and the latter is believed

to undermine the national capacities. Burnside and Dollar (1997) go on

to argue that these two are the main reasons due to which aid has no

significant effect on growth.

One of the major issues with aid is that donors impose

conditionalites. While they are not very well versed with local conditions

and requirements they ask the recipient governments to undertake

projects which are neither very feasible neither in economic terms nor

are responsive to the local context and development need. This is

particularly true in the context of disaster risk reduction, where

readymade formulas, architectured in the developed world are affected

and little effort is made to strengthen the capacity of the local population

to cope with their problems including the natural calamities which can

best be done by building upon their existing and traditional

arrangements. One of the factors which militate against the effectiveness

of disaster risk reduction strategies is the pressure on the development

agencies to disburse funds without knowing the capacities of the

implementing bodies [Seck (2007)]. For instance, during the 2000 floods


in Mozambique it was noted that some British NGOs were under

pressure from the Disaster Emergency Committee to spend money for

DRR activities on immediate basis [Hanlon (2004)].

It has been noted that the poor people in disaster-prone

developing countries are the ones who bear maximum economic losses,

and their livelihoods get affected due to their higher vulnerability to

disasters [Oxley (2013)]. The risks from the disasters can be mitigated

by adopting disaster risk reduction strategies which strengthen the

resilience of the vulnerable population. The Global Network of Civil

Society Organizations for Disaster Reduction (GNDR) (2013) supports

the Hyogo Framework for Action (HFA) which entails increased

awareness and understanding of disaster risk reduction at national and

international levels. GNDR has given recommendations for effective

DRR framework that encompasses prioritizing the poorest and

marginalized people, tackling the causes of people’s vulnerability and

improving public access to information regarding DRR.

Pakistan is a disaster prone country and is exposed to high risks

of floods, earthquakes, cyclones and landslides. Keeping this in view the

government formulated a strategy for disaster risk management known

as National Disaster Risk Management Framework (NDRMF) in 2007

[Ahmad (2013)]. The critics, however, believe that it has failed to reduce

human suffering due to bad governance, corruption, lack of political will

and overlapping responsibilities. “The government after every disaster,

perhaps in an attempt to mitigate its responsibility, cites the ‘will of God’

and punishment for wrongful deeds of the people” [Fisher (2010:552)].

On the other hand, even a still poorer country Mozambique, has

prioritized disaster risk reduction strategy for which they have a

dedicated department named as National Institute for Disaster

Management. This institute efficiently handles local operations as it has

government’s support which reflects that they value their people [UNDP

(2010)].

According to Dasgupta and Beard (2007), in order to promote

pro-poor growth, World Bank uses the term community-driven projects

that empower poor people as they have control over the development

process. Wong (2010) in his study states that “community-based

development has been criticized for its inadequate understanding of

201 Sheeba Farooq

power relationships at the local level, which thus leaves room for elite

capture”. The author further argues that these development projects tend

to fail because of weak institutional control mechanisms that create

opportunities for the local elites to gain benefit out of the opportunities.

Platteau (2004) believes that the aid given by donor agencies can only

reach the intended beneficiaries if the issue of elite capture is solved.

Wong (2010) also supports this view and suggests that a ‘co-opt-elite’

approach should be adopted which solicits cooperation of the local elites,

as they can play a vital role in the community development.

The governments of developing countries are unable to give

social protection to the poor people on long-term basis due to which the

elite get an opportunity to gain maximum shares out of the development

projects, meant for the poorest of the poor.

Natural calamities, particularly floods, disrupt the socio-

economic fabric of the society, destroy infrastructure and badly affect

the livelihoods. Populations are displaced, their regular socio-cultural

network break-down resulting into loss of ages-old social safety nets,

and their income generating activities discontinued through loss of crops,

industrial tools, cattle/poultry, shops, etc. Besides, floods make

rehabilitation of life and livelihoods even more difficult, by rendering

the farmlands uncultivable and destroying homes and businesses [Tahir,

et al. (2011)]. Mwape (2009) in his study states that floods have become

an annual event in few parts of Zambia, the floods of 2007-2008 were

the worst in terms of the amount of rainfall and the level of impact

especially on the socio-economic livelihoods of Sikaunzwe community

in Zambia. The main source of livelihood of that area was crop

production followed by trading which engaged a small proportion of the

population. The crop fields were damaged due to floods which resulted

in reduced staple crop production. This reflects their overdependence on

crop production which ultimately increased their vulnerability. Another

example is highlighted by Arnall, et al. (2013) stating that Mozambique

is a flood-prone country where more than eighty percent population

depends on agriculture which takes place in the country’s low-lying

floodplain. The worst flooding occurred in 2000 which impacted the

livelihoods of the people badly in the country. The author further

mentions about a locality Chicomo in Manhica city, where majority of


the people were involved in low-area agriculture before floods but the

percentage drastically fell from ninety nine percent to sixty five percent

after the floods due to lack of seeds, equipment and shortage of cattle for

ploughing. The severe impact of floods on the livelihoods was also

evident in the 2010 floods in Pakistan where agriculture, livestock and

fisheries suffered the highest damages, estimated at US$5.0 Billion

[World Bank (2010)]. The report produced by state that more than 80%

of the households relied on agriculture for their livelihoods. With the loss

of harvest, many poor people were unable to pay back their debts which

they had taken to buy agricultural inputs. In fact many people were

forced to do casual labour and relied on humanitarian assistance pro-

vided by the international community or richer community.

This depicts that natural disasters can lead to the long-term

impoverishment and add to the misery of the affected people who end up

losing not only their instruments of income generation, property, life-

time savings, but also get homeless and lose their social networks. This

makes them vulnerable to hazards so the need is to build their resilience

to shocks and diversify their livelihoods to make them sustainable.

Disasters not only rob people of their property, social and human

capital but also destroy their livelihoods. Effective interventions to

reduce disaster risk can lower the vulnerability of the people against

shocks and hazards [AusAid (2009)]. For instance, IFAD ensures

sustainable livelihoods by conducting vulnerability context in the flood

hit areas by incorporating DRR activities and accordingly formulate

interventions [IFAD (2013)].

The sustainable livelihood idea was first introduced by the

Brundtland Commission on Environment and Development which was

further expanded by the UN in a conference in 1992. Chambers and

Conway proposed definition of sustainable livelihood stating that “a live-

lihood comprises the capabilities, assets (including both material and

social resources) and activities required for a means of living. A lively-

hood is sustainable when it can cope with and recover from stresses and

shocks, maintain or enhance its capabilities and assets, while not under-

mining the natural resource base” [Chambers and Conway (1992: 3)].

Sustainable livelihood framework has three components;

livelihood resources, institutional processes, organizational structures

203 Sheeba Farooq

and livelihood strategies [Krantz (2001)] as shown in Figure 1.

According to this framework is used to analyze different vulnerabilities

faced by the poor people for whom they do not have the capacity to cope

with, like shocks which are at times unpredictable and traumatic in the

form of floods and epidemics. Carney (1999) has argued that livelihood

can be sustainable only if poor people have the ability to recover from

these shocks and stresses. Krantz (2001) states that people use tangible

and intangible assets to construct their livelihoods which are termed as

‘capital’ which have five types namely human capital (skills and ability

to labour), physical capital (infrastructure, shelter), natural capital

(natural resources), social capital (social relations) and financial capital

(savings and incomes) upon which people draw when pursuing different

livelihood strategies. It is argued by Scoones (1998) that institutional

processes also play a vital role as they often mediate access to livelihood

resources which make understanding of institutions important in

designing interventions which improve livelihood outcomes.

Regnier (2008) has given an example of Tamil Nadu, India,

where NGOs had restored the livelihoods of people affected by tsunami,

by adopting a sustainable development approach. The community’s main

source of livelihood was fishing, so the NGOs with the help of donors

supplied boats, motors and nets to the communities to resume their

fishing activities. The quality and diversity of the inputs have made them

capable of generating substantial revenues. In addition, they were also

given access to microcredit which made their livelihoods sustainable. A

report produced by Oxfam (2013) points out that they conduct DRR

analysis before an intervention in disaster hit area which enables them to

ensure protection of financial asset of the vulnerable people which

strengthens the performance of their livelihood programs. For instance,

heavy flooding in Beni District, Bolivia, in 2007 resulted in loss of

traditional crops and disrupted livelihood activities in twelve commu-

nities. Oxfam GB in order to make the livelihoods sustainable re-

introduced a technique that was being practiced three thousand years ago

in that particular district that comprised of modification of the landscape

such as planting of elevated seedbeds to cope with such challenges.


Figure 1. Sustainable Livelihoods Framework

Source: Krantz (2001).

Likewise DFID (1999) has also adopted SLA framework to

eliminate poverty by firstly understanding how many assets poor people

possess and then capitalizing on those assets as they are interconnected.

There are different factors that either hinder or enhance the livelihood

opportunities of poor people such as access to different assets, culture,

norms and institutional processes which should ultimately be responsive

to the needs of the poor. For instance, provision of microcredit to the

poor people or government providing social safety nets can increase their

sources of income. The more sources of income people have the more

strength they have to cope with and recover from shocks and stresses

which make their livelihood sustainable [Chambers and Conway

(1992)].

This study will be adopting this holistic framework and examine

as to how were the people affected by 2010 floods in Pakistan and what

did the donors do in order to restore their social, physical, human, natural

and financial assets which are the capital endowments required to create

livelihoods, but the main focus will be on financial capital.

The Citizen Damage Compensation Program (CDCP) emerged

to be the single major initiative launched by the Government of Pakistan

to support the rehabilitation of the 2010 flood affectees. The programme

was undertaken in two phases. During Phase I the Federal Government

VULNERABILITY

CONTEXT

• SHOCKS

• TRENDS

• SEASONALITY

STRUCTURES • Levels of government • Private Sector • laws

• Policies

• Culture

• Institutions

PROCESSES

TRANSFORMING

STRUCTURES &

PROCESSES

LIVELIHOOD STRATEGIES

LIVELIHOOD

OUTCOMES

• More income • Increased Wellbeing • Reduced vulnerability • More sustainable use of NR base

i n o r d e r t o a c h i e v e

S

H

N

P F

LIVELIHOOD ASSETS

Influence & access

Key H = Human Capital S= Social Capital N = Natural Capital P = Physical Capital F = Financial Capital

205 Sheeba Farooq

in collaboration with the Provincial Governments embarked upon the

ambitious programme of unconditional cash transfers to flood affectees

to help them restore their livelihoods. A total fund of US$400 million

with the contribution of US$200 million by Federal Government and

US$200 million by Provincial Governments was arranged. It was agreed

to provide cash assistance of US$230 approximately to each affected

family. The Provincial Disaster Management Authorities (PDMAs) were

required to identify flood affected areas in each province; National

Database and Registration Authority (NADRA) an arm of Ministry of

Interior, GOP, was to verify the lists of family heads within the affected

areas and commercial banks were to make the payments to the

beneficiaries through a debit card termed as the “Watan Card”.

According to a study by Emergency Relief Cell (2012) of the Cabinet

Division, GOP, within three months of the programs initiation, 1.4

million families were registered and paid US$230 each.

Some studies for instance that of Hunt, et al. (2011), however,

are skeptical about the effectiveness and performance of this program

and observes that the pre-condition to become a beneficiary of CDCP

was to have a valid CNIC (Computerized National Identity Card), which

most of the poor and vulnerable did not have. Moreover, people who did

not update their details at NADRA database and who were living in small

communities, far away from the main villages and NADRA offices were

filtered out. The study also highlighted the factors such as lack of

communication between policy makers and implementers, lengthy

procedures and capacity issues at the administration level, due to which

the eligible beneficiaries were unable to get access to the CDCP cash

transfer. The authors also point out that out of 100 eligible family heads

only 43 received Watan Cards on average in Sindh and Balochistan

provinces which shows that many eligible people were left out and the

grievance mechanism did not function properly to give justice to the

aggrieved.

On the contrary, the National Database and Registration

Authority (NADRA) claims that total appeals filed with grievance

redressal cell were 385,010 out of which 39,105 had been resolved and

345,905 were under scrutiny as of May, 2012 (NADRA, 2012). It has

also been claimed that “Ninety percent genuine flood affectees had


received Watan Cards (Government functionary (personal comment) 26

June 2013)”. The Grievance Redressal Cell was set up right after the

implementation of Phase I so that the beneficiaries who had been left out

could be facilitated. Its purpose was to scrutinize beneficiaries’ requests

and complaints vis-à-vis the payments and provide a mechanism for

social accountability of the program.

It was because of the overall satisfactory results of CDCP Phase

I which encouraged the development partners to collaborate with the

GOP for CDCP Phase II. Accordingly, World Bank pledged to

contribute US$125 million, DFID US$100 million, USAID US$190

million and Government of Italian Republic US$65 million. This total of

US$480 million with a contribution of US$100 million by GOP was

agreed to be disbursed to 1.21 million households in two equal tranches

to help them rebuild their lives [World Bank (2011); GOP (2012)]. In

order to make the implementation of Phase II effective the World Bank

had undertaken rapid evaluation of Phase I in January 2011 which ascer-

tained and highlight the weak areas of the first phase. For instance, some

of the flood affected families’ names were missing in the beneficiaries

list. Moreover, events of malpractices were noticed such as unsystematic

handling of complaints and demands of bribes by the local officials who

were conducting the survey to expedite the complaints process. But

overall the evaluation mission considered the progress of program to be

satisfactory as 900,000 beneficiary households had been given Watan

Card [World Bank (2011); GOP (2012); World Bank (2012)].

However, to further improve the programme and to maximize its

effectiveness at the suggestion of the World Bank, a refined targeting

mechanism was adopted whereby better-off families were filtered out

and legitimate vulnerable beneficiaries such as female and disabled

headed households were included in Phase II [ERC (2012)]. In addition,

it was decided that the selection criteria used as a proxy for eligibility,

should be based on the beneficiary’s damaged house. Each province

conducted a house damage assessment survey which identified the Heads

of the Household of the damaged houses and PDMA shared the data with

NADRA for verification which was further sent to Third Party

Verification Firm for validation to ensure maximum accuracy and

transparency [ERC (2011)]. A similar study carried out by the World

207 Sheeba Farooq

Bank (2011) further emphasized that the payment to beneficiaries under

CDCP be made conditional on having a valid CNIC. The payment cycle

was of three months between each installment. The beneficiaries were

expected to prudently use these funds to recapitalize their assets, restore

their livelihoods and repair their houses.

Accordingly the CDCP, particularly the Phase II, which was

undertaken primarily with the support of the development partners, did

contribute towards restoration of livelihoods, as people in most cases

utilized the cash assistance provided to them for buying cattle, poultry,

basic tools for cottage industry and fertilizer for reactivating their

agricultural activities. It also helped them to repair their dwellings and to

render them livable. Most of the people moved back to their places of

residence and restarted their normal socio-economic activities in their

familiar surroundings with the help of Watan Card assistance.

In summation, international aid and donor’s assistance has

actually led to improvement in the well-being of the poor in the

developing world and has saved lives at the times of natural calamities

and disasters. However, it has not led to a process of sustainable growth

and countries around the developing world continue to depend upon

financial and technical support from multilateral and bilateral donors.

Whether these are purely economic crisis such as widening budget

deficits, depleting foreign exchange reserves or the nature inflicted

catastrophes like tsunami, earthquake and floods, the governments keep

looking towards industrial west for support.

In spite of being repeatedly exposed to natural calamities and

having seen thousands of lives lost and their infrastructures destroyed

the developing countries still continue to remain reactive in their

approach rather than pro-active to prepare for the natural disasters. Some

countries have established organizations such as National Disaster

Management Authority (NDMA) and Provincial Disaster Management

Authority (PDMA) in Pakistan, but they are rather at an embryonic stage,

without adequate technical expertise and know how. They badly lack

necessary equipments, technical facilities and necessary training.

During 2010 floods Pakistan was in dire need of help from the

international community to cope with the catastrophe. The international

community came forward with committing an amount of US$1.371


billion against some one hundred and forty four projects as already stated

above, covering a wide spectrum from provision of medical health to

food items and shelter. But it should be noted that the main instrument

for the restoration of livelihoods was by Watan Card under CDCP. The

Watan Card Scheme was divided in two phases, the purpose of the first

phase was to give immediate relief to flood affectees to fulfill their

pressing needs and the aim of cash assistance given in Phase II was to

restore the livelihoods of the flood affected population. There were some

issues and snags faced by the program, as has been mentioned above,

and there have been criticism of the same, but if we look at the bigger

picture Watan Card has greatly contributed in the restoration of poor

people’s livelihoods, as most of the people utilized the cash for

purchasing agricultural inputs, livestock, hand looms and restoring their

agricultural lands, and have started to generate their income which is

being further utilized in sending their children back to schools and

ultimately spending on health and nutrition.

In view of the above analysis, this study adopts the following

research objectives in order to ascertain the level and effectiveness of the

support extended by the donors for the rehabilitation of 2010 flood

affected population.

3. METHODOLOGY

3.1. Research Approach

This study is based on a case-study of a village named Misri

Banda in Nowshera district in Khyber Pakhtunkhwa (KP) Pakistan, to

examine the impact of 2010 floods and the role of donors in bringing the

people back to pre-flood conditions, with a focus on restoration of their

livelihoods. The method of case study was used because in this approach

different methods are combined with the purpose of illuminating a case

from different angles [Johansson (2003)]. It is a tool to find out the true

picture which is mostly not depicted in the reports and articles.

In order to carry out the case study a qualitative research

approach was used shown in Table 1, as it allows the evaluator to study

the issues in depth and data collection is not restricted by predetermined

categories [Patton (2002)]. The qualitative instruments administered

209 Sheeba Farooq

included focus group discussions (FGDs) with bene-ficiaries, in-depth

interviews with non-beneficiaries, semi-structured interviews with the

beneficiaries, officials of donor agencies and the community leaders to

validate the data collected through questionnaires. Focus group

discussions were carried out to make in-depth analysis of the experience

of beneficiaries. The groups were identified through community support

[Chaudhry (2013)]. The sample included normal beneficiaries, female

headed-household and vulnerable people ensuring equal inclusion of

males and females.

Table 1. Research Methods Used in Study

Method Number of Exercises Exercises Conducted With

1.Semi-

structured

interviews

-Seventy interviews were conducted

with the flood affected people.

-Ten interviews were conducted

with the government officials.

-Ten interviews were conducted

with the development functionaries.

-Meetings with Government of

Pakistan functionaries dealing with

development assistance.

-Meetings with relevant functionaries

of donor organizations operating in

Pakistan.

-Assessing affected population’s

perception in this regard.

2.Focus

Group

Discussion

-Ten focus group discussions were

carried out to find out the views of

people on cash assistance.

-The discussions were conducted with

f female-headed households, males and

vvulnerable people.

The sampling technique used was Simple Random Sampling,

because each element has an equal probability of being chosen as the

subject [Sekaran and Bougie (2010)]. The first step was to determine an

appropriate sample of beneficiaries which could represent the target

population. Based on the sampling technique the total sample size

selected was 80 beneficiaries, out of which seventy were interviewed,

comprising of 35 men and 35 women, from different households and 10

focus group discussions were held with equal representation from the

males and females of that community. Four specific situations were

analyzed in-depth to document respondent’s experience regarding the

effectiveness of the donor’s assistance.


3.2. Data Collection

The data were collected over a period of three to six months in

2013. The identification of village and data collection facilitation was

provided by the Deputy Commissioner, Nowshera. The researcher

undertook the interviews along with the support from local admin-

istration. The research used both primary and secondary sources of data.

Primary data, refer to information obtained first-hand on the variables of

interest for the specific purpose of the study [Sekaran and Bougie

(2010)]. Examples of sources of primary data are interviews, admin-

istered questionnaires, focus groups in which people from different age

groups are invited to give their opinion on specific issues, video

conferencing and the internet if the questionnaire is administered over it.

Accordingly, focus group discussions and semi-structured interviews

were held in 2013 with the respondents identified by the focal person to

get villagers’ opinion on the specific issues to determine the level of their

satisfaction and to explore the issues pertaining to that context. The other

source of data used is secondary data, which refers to information

gathered from company archives or records, websites, periodicals,

census data, statistical abstracts and the media [Sekaran and Bougie

(2010)]. But in this research the data were collected by the existing

literature which includes government and donor publications, reports

published by the NGOs, newspapers and relevant articles.

Ethical clearance was obtained prior to the commencement of

research and permission of interviewees was also sought before inter-

views were held. All information obtained from the interviews was

treated as confidential. The interviews generated a wide variety of

opinions on how impact of floods could have been mitigated and as to

how successful the rehabilitation process has been.

The findings provide substantial representation of affectees’

experiences which reflect the voice of the majority of people living in

that particular community.

4. RESULTS AND ANALYSIS

This section addresses the three sub-objectives of the study,

which seek to explore as to how the people were impacted by 2010

211 Sheeba Farooq

floods, with particular reference to Nowshera district in Khyber

Pakhtunkuwa province in Pakistan, and to find out the value and extent

of donor’s contribution towards restoration of the livelihoods of the flood

affectees, particularly for building their financial capital. The results and

analysis of the interviews and focus group discussions held with the

flood affectees, functionaries of multilateral as well as bilateral donor

organization based in Pakistan, and government officials are presented

throughout this section.

While the development partners were initially reluctant to

participate in CDCP, they joined it at a later stage for Phase II with a

combined contribution of US$480 million. Out of the total US$480

million, USAID contributed US$190 million, World Bank US$125

million, DFID US$100 million and Government of Italy US$65 million.

The GOP also made a contribution equal to US$100 million [World

Bank (2011)].

In the meanwhile, the bilateral/multilateral donors and aid

agencies continued to provide support and assistance to flood affectees

through provision of food items, blankets and health services which

greatly helped the early recovery efforts. The main window for

restoration of livelihoods of flood affectees, however, was the CDCP

which provided a sum of US$200 and $400 to each flood affectee during

Phases I and II, respectively.

Our literature review as well as field study amply manifests that

the donors made substantial contribution towards the rehabilitation of

flood affectees, particularly for CDCP Phase II, which was the main

instrument to restore the livelihoods of the flood affectees. While the

cash assistance provided in Phase I was predominantly used by people

to meet their immediate needs, it was the assistance received during

Phase II which was utilized for purchase and restoration of basic

instruments of income generation. People purchased livestock and

fertilizers with the money, restored their agricultural lands, started small

businesses, and repaired their houses to render them livable. This finding

is fully substantiated by our field study of Nowshera District.


4.1. Case Study of Misri Banda Village in Nowshera District

Nowshera District is situated at south east of provincial capital Peshawar

in Khyber Pakhtunkawa (KP) province, where Kabul River joins River

Indus. The total area of the district is 1748 sq km with a population of

1.1 million [NADRA (2012)]. The main source of income of the people

is agriculture. Nowshera was one of the worst affected districts in KP,

the floods wrecked the infrastructure, sweeping away houses, roads,

bridges, electricity network and agricultural land [UNDP (2013)]. Out of

the total population, almost ninety percent was displaced from its

original place of residence, losing their livestock, cattle, small businesses

and cottage industry [PDMA (2010)]. The destruction caused by the

floods can be well assessed from Table 2.

The devastation was caused in all sections of life, livelihoods

both agriculture and non-agriculture were directly or indirectly affected

as depicted in Table 2 [UNDP (2013)].

Table 2. Summary of Losses/Damages

District Total

Population Dead Injured

Total

Population

Displaced

H.H

Affected

Educational

Facilities

Health

Facilities

Nowshera 1.2 Million 167 10 350,336 71,403 134 11

Source: The data have been taken by the PDMA report on Summary of Losses/ Damages in 2010.

4.1.1. Misri Banda

Misri Banda is a small village in Nowshera District with a

population of 12,303 and literacy rate of about 31% [Pakistan Bureau of

Statistics (2013)]. It has a high poverty rate with people meeting two

ends with difficulties and residing in mud houses. Most of the people are

engaged in agricultural activities though some earn their livelihood

through livestock, cottage industry, small enterprise, and masonry work.

Being one of the poorest communities worst hit by the 2010 floods with

a major part of the population displaced and severely affected, Misri

Banda was considered to be the most appropriate place for a case study

for this research and thus selected.

213 Sheeba Farooq

4.2. Impact of Floods on Misri Banda

Interviews conducted at Misri Banda indicate that 94% of the

population was affected by 2010 floods and as a result, a number of

families lost one or two members, 35% had their properties inundated

and 94% lost their instruments of income generating activities. The

economy of Misri Banda is agriculture based with 58% engaged in

agriculture as their main source of livelihood while 80% depend on

livestock and poultry, 13% run small enterprises and 21% works as daily

wagers/labourers [NADRA (2013)]. Eighteen percent2 of the people

owned 1-2 hectares of agricultural land before floods from which they

earned £100 to £120 per month, while farm-tenants and other daily

workers earned approximately £2-3 a day. The average monthly income

of the people engaged in small enterprises and cottage industry stood at

around £100, a rather meager amount to fulfill family needs. The

residents of Misri Banda were thus surviving at a subsistence level with

minimum possible facilities of life even before the floods, which too,

were taken away from them by this unexpected catastrophe.

One of the residents of Misri Banda interviewed regarding the

impact of floods stated that “the flood was a dreadful incident, it struck

the village at night when the electricity was out, people were screaming

for help and it seemed that nobody will be able to survive. People were

running for their lives leaving behind their entire assets made out of life

time earnings and important documents such as identity cards. There was

chaos everywhere and with a blink of an eye the flood washed away

houses, bridges, infrastructure and people were helpless”.

4.3. Restoration of Livelihoods of the Flood Affected People:

Role of Donors

As in other flood-affected areas the government undertook relief

and recovery efforts in Misri Banda as well. People were rescued out of

heavy flowing water with the help of helicopters and boats and over 70%

of the population was moved to IDP camps where they were provided

with food items, tents, medicines, etc.

2 Tabulated information is available with the author.


After the floods, to rehabilitate the livelihoods of the people the

government initiated the cash transfer program under CDCP Phase II. As

has been stated above, CDCP Phase I was for early recovery while CDCP

Phase II (Watan Card Scheme) was for rehabilitation of the flood

affectees, where the development partners joined hands with a

contribution of US$ 480 million.

During the field visit a large number of people were encountered

who had benefitted from the Watan Card Scheme but on the other hand,

there were people who were not contented with this compensation. When

asked from the respondents about the utilization of cash; 30% reported

that the cash received from Watan Card was utilized in purchasing

livestock, 18% restored their agricultural land, 12% repaired houses,

10% started small business, 12% met their daily expenses, 8% used it for

educational purpose and 4% used it on health care. It should be noted

that people had spent part of the money in repairing their houses, part of

the money in purchasing instruments of production and part of the money

on meeting daily expenses. Critics argue that majority of the bene-

ficiaries of Watan Card had spent the amount on their immediate

consumption needs, particularly food, clothing and healthcare rather than

on restoring their livelihoods [Hunt, et al. (2011)]. Even if this was true

in some cases, most of the families did utilize the amount for re-

activation of their income generating activities. Besides, the amount

spent for immediate consumption too led towards the restoration of their

lives and was unavoidable.

Most interviews indicate that the cash assistance was enough to

restore small enterprises such as a shop or cottage industry or to purchase

poultry and basic gadgets such as a hand loom as is evident from Figure

2. For instance, 82% respondents agreed that they were able to restore

their cottage industry but 18% were unable to do so as they had spent the

money elsewhere. On the other hand, 74% respondents said that the

amount was inadequate to purchase livestock; reason being that a buffalo

costs about £300-350 and the amount that was given was £258

approximately in Phase II which too was given in two tranches. But the

twenty six percent who were satisfied were those who bought poultry or

a small goat. When asked the respondents about the difficulties they

faced in getting the compensation; ten percent replied that the authorities

215 Sheeba Farooq

were not accessible, 24% replied that they were not helpful, they were

rude, and indifferent, ten percent reported that they needed contacts, as

the field functionaries tried to fleece people who had no references, and

nine percent reported that officials wanted bribes. On the other hand,

36% respondents stated that they were not aware of any cash assistance

or compensation available but 64% respondents reported that the local

mobilizers, national television broadcast and radio disseminated the

information among flood-affectees. Findings suggest that the people who

could not receive compensation were those who had either lost necessary

documents like national identity card to establish their eligibility or did

not have the transportation to reach the facilitation centre.

In addition to the regular multilateral and bilateral donors such as

the World Bank, USAID, DFID and Italian Republic which made

substantive contribution towards cash assistance (Watan Card Phase II),

a number of other countries and international humanitarian organizations

also came forward with assistance in cash and kind to assist the flood

affec-tees. For instance, Solidar Switzerland supported the displaced

people by building their capacities and restoring livelihoods in Misri

Banda village. Two hundred and five women received training in

embroidery and dress designing and in addition a number of villagers

were given tool kits and their links were developed with the market so

that they can sell their products and earn their livelihood [Solidar

(2013)].

Likewise, IOM’s initiation of Cash for Work Scheme in Misri

Banda, in which the flood affectees were given training in house building

and construction so that they can become productive and earn cash and

may also construct their own houses, was a win-win situation for them.

Another International NGO Action against Hunger, developed

new water supply schemes that provided safe drinking water to the Misri

Banda community [ACF International, (2012)]. Hence, number of

international organizations played a vital role in the rehabilitation of the

flood-victims.

Results show that 86% people received basic necessities from the

donors in relief phase which comprised of bedding, cooking utensils,

medicines, food items and tents. The other 14% respondents informed

that they did not receive immediate assistance from the donors or


government agencies as they lived in far flung areas that were not

accessible.

Figure 2. Utilization of Cash for Restoration of Livelihood

Source: Based on the Author’s calculation.

When specifically enquired about the support provided by the

international community/donors while some respondents expressed their

dissatisfaction about 69% acknowledged the presence of international

community and appreciated their contribution towards socio-economic

rehabilitation of the flood-affectees.

4.4. Discussion of Results

The field study of Misri Banda, and the focus group discussions

as well as one-to-one interviews, however, bring us to a safe conclusion

that almost 70% of the population received Watan Card which helped

them in resettling and restoring their financial capital. Out of total

respondents nine percent claimed that they did not receive the Watan

Card due to the non-availability or loss of documents like national

identity cards during floods, which was necessary for proving eligibility

for Watan Card. They were unable to get the documents made due to

30%

18%12%

10%

12%

8%4%

30% purchasedlivestock

18% restoredagricultural land

12% repaired houses

10% started smallbusiness

12% met daily expenses

8% educational purpose

4% health care

217 Sheeba Farooq

inaccessibility to the offices responsible for issuing documents as they

were living in far flung areas. Out of the total respondents ten percent of

the population tried to become illegal beneficiaries of the cash assistance.

It was reported about a boy, Raja, who had left his job in the city and

came back to his village hoping that he would receive Watan Card,

although his house had not been damaged. This incident made the

researcher curious to probe further and a few young boys in their twenties

were interviewed. It was a shocking revelation that hundreds of boys left

their jobs in cities hoping that they will become claimants of Watan Card

back in their villages. These jobs were their only source of income and

they left them even though there was no surety whether they will get the

Watan Card amount or not. This in fact is a negative aspect of cash

assistance which creates dependency syndrome among the people.

Semple (2011) also refers to this in his study, stating that aid had a

negative impact on majority of the people during 2010 floods in

Pakistan, as they became dependent on aid and did not want to become

self-sufficient.

During the interview with an INGO officer it was found out that

“46000 complaints had been registered by February, 2013 in Nowshera

for not receiving Watan Card and almost ninety percent of them had been

resolved by now” (INGO officer3 (personal comment) 5 June, 2013).

The findings also reveal that in a very few cases the whole

amount received through Watan Card was spent on one particular

activity. Almost everywhere it was utilized for number of needs though

major amounts were spent to repair houses or purchase cattle, hand

looms, goods for sale at a shop or reactivation of a commercial activity

disrupted by the floods. It may however be noted that before the cash

assistance, flood affectees were borrowing money to fulfill their

immediate needs either from friends, relatives or shopkeepers [IOM

(2012)]. That was the reason the government came up with the idea of

giving unconditional cash grant to the flood-affected population. A

report produced by International Office of Migration stated that

“international evidence suggests that cash grants allow the recipients the

3 Interview transcript is available with the author.


flexibility of choosing where to put their resources based on their specific

conditions and priorities” [IOM (2012: 24)].

The Watan Card helped flood affectees to restore their liveli-

hoods and made them capable to revive their income generating

activities, and to rebuild their lost assets. According to some respondents,

the amount of Watan Card was not sufficient but as a matter of fact, the

assistance did help them to revive their incomes to pre-flood level as

shown in Figure 3. As has been stated earlier most of the people living

in Misri Banda were earning £95-100 per month before floods, the

amount of Watan Card which was almost £258 was thus a very

substantive support.

Figure 3. Respondent’s Views on Cash Assistance

Source: Based on the Author’s calculation.

Empirical evidence also shows instances of Elite Capture during

the 2010 floods in Pakistan whereby the local politicians and feudal lords

managed diversion of funding to their constituencies to strengthen and

develop their vote banks. It was informed by a senior officer of Ministry

of Finance and Economic Affairs during an interview that “a famous lady

politician, the then Minister of State for Economic Affair Division, had

called a meeting of Pakistan Poverty alleviation Fund (PPAF) whose

mandate is to reduce poverty around the country. PPAF were advised to

discontinue other projects and rush teams to her constituency to establish

camp offices and provide relief to the flood affectees to strengthen her

70%11%

9%

10%

70% satisfied withWatan Card

11% not satisfied withWatan Card

9% non-availability ofdocuments

10% illegal beneficiaries

219 Sheeba Farooq

vote bank [Anon. (personal comment) (28 June, 2013)]”. This incident

was also confirmed by the senior management of PPAF. It was also

endorsed by an NGO functionary4 who did extensive research in the

context of rehabilitation of flood affectees and studied the phenomenon

of elite capture in depth. According to him “a good amount of cash

assistance was diverted to the lady politician’s constituency / home

district while the needy did not get cash in some parts of Muzaffargarh,

South Punjab. She had asked most of the ambassadors to give everything

to her constituency [NGO functionary (personal comment) (7 June,

2013)]”. Another powerful feudal family in Shikarpur district of Sindh,

the Bajranis managed to get hold of most of the assistance meant for

rehabilitation of flood affectees of that area, on the pretext that they will

directly be distributing it to the deserving people.

An interesting incident was quoted in a study by Semple (2011)

which is another example of elite control. He states that 100%

destruction was reported in Sabzujat Union Council in the district

Muzaffargarh, South Punjab, and the whole council was submerged in

water but as per his research ‘hardly anybody there could qualify for

Watan Card as they had voted against the Minister for Economic Affairs,

during the last national elections [Semple (2011: 77)]. It is the same lady

Minister who has been referred above. Likewise there were also reports

confirmed by a number of media and civil society observers that some

powerful feudal lords and politicians such as Jamalis, a very big clan in

the border areas of Boluchistan and Sindh, had the rivers and canal

embankments fractured at specific points to reduce the flow of water

before reaching their lands [Ahmed (2013)]. This did save the crops and

lands of a few wealthy and mighty but aggravated human miseries, as

the specific points where the water was allowed to outflow were densely

populated areas with small villages, which as a result were exposed to

unexpected levels of floods.

In spite of the misuses of the distribution of Watan cards, the

process in general has been quite fair as claimed by the World Bank and

NADRA and confirmed through third party validation in the field. One

million beneficiaries received assistance through Watan Card and the

4 Interview transcript is available with the author.


World Bank declared it as the flagship programme of GOP and a symbol

of one of the largest humanitarian cash transfer programs. Moreover, a

survey conducted by FAO has emphasized that “Watan card had the

broadest coverage of any kind of intervention in the early recovery

period and it has been recognized as a model that other disaster stricken-

countries might consider emulating” [FAO, (2012: 7)].

To the extent of immediate rehabilitation of livelihoods, the

foreign aid received for the 2010 flood affectees has thus been fairly

effective. However, if assessed from the perspective of Disaster Risk

Reduction, it does not seem to have made much difference. The people

as well as their socio-economic infrastructure remain as vulnerable to

natural calamities as before. Their dwellings and commercial set-ups

have not improved to better resist future floods or similar disasters nor

have any measures been taken to reduce the destructive impact of floods,

by strengthening river banks or managing the flow of torrential waters.

For this, as has been discussed above perhaps the political expediencies

and the element of Elite Capture is more to blame, whereby the proposed

plan of development partners to ‘Build Back Better’ was rejected and

instead the cash assistance was opted for.

5. CONCLUSION

Pakistan was hit by a catastrophic flood in 2010, which affected

over 20 million people in terms of loss of lives, shelter and livelihoods

and inundated an area stretching over 100,000 sq km destroying all kind

of infrastructure and facilities. The country was hardly prepared to meet

the disaster and its after effects. However, rescue and relief efforts were

undertaken with the active support of the international community. Yet

the more daunting task was to restore the damaged infrastructure and to

rehabilitate the displaced population. For this the Government of

Pakistan in collaboration with the development partners undertook

CDCP Phase II as has been discussed in detail in this paper.

The GOP and development partners claim that the CDCP Phase

II has been a great success and is one of the best cash transfer

programmes ever under taken in the world for disaster effected

population. The programme has come to finality and most of the flood

221 Sheeba Farooq

affectees have resumed their regular socio-economic activities, at their

places of pre-floods residence.

However, there have been instances of mismanagement and

misuse of the funds during the process for instance; some politicians

manipulated the situation to divert some funds to their constituencies to

strengthen their vote banks, even though those areas were not hit by the

floods. Similarly, some field functionaries of the government managed

to exploit the illiterate population by demanding share in the cash

assistance to which they were entitled. There were also instances of field

functionaries refusing to accept even genuine documents such as identity

cards and authenticated payment of assistance only at the receipt of

bribes. In some cases the badly damaged infrastructure and inadequate

data also militated against the funds reaching to the deserving people.

Some cases of the affected population trying to get more than their

entitlement by offering bribes and presenting fake documents were also

noted. Some people who had not been affected by the floods also tried to

make hay while the sun shines. In spite of all the above issues CDCP has

attained its objective to rehabilitate the flood affected population to a

good extent as per the claim of the GOP, the World Bank and other

development partners. The interviews conducted for this research and the

field study undertaken in Misri Banda, Nowshera District, also support

this claim.

The ground reality that majority of people have moved back to

their native towns and original places of residence, have resumed

professional/income generating activities, their children have started

going to schools/madrasas and their social lives have more or less been

brought back to pre-flood mode; bears ample witness to the fact that the

efforts by the GOP and the donors for the rehabilitation of the 2010 flood

affectees have been fairly successful.

For future the GOP needs to put in place necessary mechanism to

meet such disasters. Pakistan is a flood-prone country and the monsoon

rains in 2013 have again been very heavy, resulting into loss of life and

destruction of infrastructure. The Daily News in its editorial on August

6th, 2013 reads that “100 people have died due to the latest monsoon

which has become an annual lament but unfortunately the country is

never prepared to deal with the destruction caused by floods”. The report


further highlights that ‘the National Disaster Management Authority

(NDMA) has, for some reason, only been allocated Rs.180 million (£1.1

million) for the coming fiscal year, even though it ended up needing Rs.5

billion (£33 million approximately) last year and that the NDMA needs

to be proactive and formulate a strategy rather than wait for the disaster

to hit the country (The News, (2013: 7)].

Thus, the GOP must formulate long-term and effective strategy

and allocate adequate funds for Disaster Risk Reduction. NDMA and

DMAs need to be fully equipped with modern machinery/ technology

and its functionaries must be provided training perhaps with the help of

development partners to meet future eventualities. The government must

develop and put in place a comprehensive policy framework and

required wherewithal in which the development partners can support it

through advice and assistance.

On their part, the development partners and international

community, to ensure effective and best utilization of their aid, should

insist that the governments in the disaster prone countries improve

capacities and infrastructure to minimize the destruction and damages by

the natural calamities. They should also support such initiatives which

enhance self-sufficiency of the populations to meet their economic needs

and to cope with their natural disasters rather than be dependent on cash

assistance and government support.

The scope of the study was towards a specific area i.e. impact of

floods on livelihoods but future research can be conducted to find out

how to bring in behaviour change in the people who reside in flood-prone

areas to build their resilience and ultimately transform the society. This

assessment will give an explicit direction to the government to come up

with more effective programs in the future. The study’s limitation were

the non-availability of extensive data, comparative analysis of various

districts. Also, very few organizations had studied the impact assessment

of Watan Cards.

6. POLICY IMPLICATIONS

Based on the research, several policy implications are formulated are

stated below:

223 Sheeba Farooq

1. The cash assistance policy was good on short-term basis as it

facilitated rehabilitation of the flood affectees. However, it did not

contribute enough on long-term basis such as to prepare the

population for similar disasters in the future. The best way could

have been to provide support for immediate rehabilitation but

enjoining it with long term planning and undertaking, what the World

Bank called ‘Build Back Better’. The reconstruction, particularly of

infrastructure essential for maintaining the income generating

activities of the flood-prone areas should have been done at a level

that it could resist damage and destruction by future floods.

2. While the 2010 flood affectees have more or less been rehabilitated

and have resumed their normal lives, the broader issue of reducing

their vulnerability to floods and other natural calamities remains

unaddressed. People have moved back to their original places of

residence but these exactly are the areas which have maximum

exposure to floods and similar disasters. The monsoon rains in 2013

had again been unexpectedly very heavy and triggered heavy floods

resulting into loses of lives, properties and infrastructure. As per the

report of The News dated 4th August, 2013 “in one day the flash

floods have cost over sixty lives, disrupted communication network

in most of KP, Azad Jammu Kashmir and Baluchistan and caused

heavy damage to infrastructure”. In yet another report the newspaper

quoted KP Minister for Information and PDMA Chief saying that

“efforts are on to minimize flood damage”. Now this is what has been

referred as a reactive approach in this study. The Federal and

Provincial governments in Pakistan need to come out of this mode

by adopting a pro-active approach which will be a two-pronged

strategy. For instance, risk-assessment should be conducted prior to

emergency planning to ensure that organizations possess sufficient

supplies to respond to emergencies. Dedicated structures should be

in place for imparting technical training to staff for crises

preparedness [OECD (2013)].

3. Instead of making efforts with a reactive approach, to minimize the

damages after the calamity has occurred, there must be standing

operating procedures or SOPs available to relevant agencies such as

NDMA, PDMAs and district governments, who should remain


prepared to meet an exigency. The Government should formulate a

National Flood Management Policy as in Indonesia whereby their

government has organized disaster task force to coordinate the

emergency management. A national body provides central

coordination, with support from technical ministries (Studari, 2004).

Secondly, the government should develop relocation plans for flood-

prone areas, build rescue stations and invest in human capital to

overcome emergencies.

4. According to another report in The News, 8th August, 2013 World

Bank has issued a policy note titled “Managing Natural Disaster” to

Pakistan to improve disaster management by creating city emergency

centers in major urban centers. The policy note further recommends

“establishing linkages between city emergency centers and current

disaster management structures at federal, provincial and district

levels to ensure timely response in case of any natural disasters such

as floods, earthquakes, excess rainfalls, cyclones, tsunamis etc.” The

policy note also states that government should ensure systematic

methods for information gathering regarding disaster risk. The paper

further highlights that ‘the capacity of municipal governments is

quite limited due to lack of communication systems, equipment’s and

technical capacity’. It thus advises the government to ensure physical

resilience by making new infrastructure flood-prone and abiding by

the building codes through improved supervision of new

construction taking place [Haider (2013: 15)]. The government

policy, therefore, must ensure that the issues highlighted by the

World Bank, which have been time and again flashed by other

development partners, civil society and the media, are addressed on

war footing. The GOP with the help of development partners also

need to institute Rapid Response Force well-equipped with technical

skills and necessary tools. The government should conduct damage

assessment in collaboration with donors for targeting flood affectees

effectively. Secondly, the government can establish an emergency

contingency fund in collaboration with donors which will help in

restoring their livelihoods and reconstructing homes [World Bank

(2010)].

225 Sheeba Farooq

5. The GOP and development partners need to make comprehensive

impact assessment of the cash assistance programme, which is

reportedly under process to identify the gaps and weaknesses, if any

so that they are properly addressed in such future situations.

6. While some cash assistance to meet immediate needs may be

necessary, donors should as a matter of policy prevail upon

governments in the developing countries to institute projects with

donor funding, which productively employ disaster hit populations

leading towards their income generation on the one hand and skill

development for maintaining livelihoods on self-sufficient basis on

the other hand.

The government should be focusing on developing proper

relocation plans for flood-prone areas, build rescue stations, install

earning warning system and invest in reconstruction of the infrastructure.

Every constituency should be allocated funds for rehabilitation during

emergencies to curb elite capture. The government should conduct risk

assessment and ensure that relevant organizations have sufficient

mechanism in place to deal with disasters. The government should be

working in collaboration with donors to devise disaster risk reduction

strategies which will help in minimizing the damages caused by natural

disasters.

REFERENCES

ACF International (2012) Humanitarian Support to Conflict and Flood-

Affected Populations in Pakistan. Islamabad: Pakistan.

Ahmad, S. (2013) The Unbridled Invitations to Floods In Pakistan.

Viewpoint. 12 July. Assessed on 10 May 2013: http://www.view

pointonline.net/the-unbridled-invitations-to-floods-inpakistan.ht

ml.

Ahmed, Z. (2013) Disaster Risks and Disaster Management Policies and

Practices in Pakistan-A Critical Analysis of Disaster Management

Act 2010 of Pakistan. International Journal of Disaster Risk

Reduction, 4, 15-20.


Arnall, A., D. S Thomas, C. Twyman and D. Liverman (2013) Flooding,

Resettlement, and Change in Livelihoods: Evidence From Rural

Mozambique. Disasters, 37:3, 468-488.

AusAid (2009) Investing in a Safer Future: A Disaster Risk Reduction

Policy for the Australian Aid Program. Canberra: Australian

Agency for International Development.

Baubion, C. (2013) OECD Risk Management: Strategic Crisis

Management. OECD Working Papers on Public Governance, 23,

0-1.

Brooker, J. (2011) One Year On 2011: Pakistan Floods. Islamabad:

United Nations.

Burnside, C. and D. Dollar (1997) Aid Spurs Growth-In a Sound Policy

Environment. Finance and Development. 34:4, 4-7.

Carney, D. (1999) Approaches to Sustainable Livelihoods for the Rural

Poor. Overseas Development Institute.

Chambers, R. and G. Conway (1992) Sustainable Rural Livelihoods:

Practical concepts for the 21st century. IDS Discussion Paper 296,

Brighton: IDS.

Chaudhry, A. I. (2013) Operational Audit Household Survey Report:

Citizens’ Damage Compensation Program. Deloitte Islamabad:

Pakistan.

Dasgupta, A. and V. A. Beard (2007) Community Driven Development,

Collective Action And Elite Capture in Indonesia. Development

and Change. 38:2, 229-249.

DFID, U. K. (1999) Sustainable Livelihoods Guidance Sheets. London:

DFID.

Emergency Relief Cell (2011) Citizen’s Damage Compensation

Program: Operational Manual. Cabinet Division, Islamabad,

Government of Pakistan.

Emergency Relief Cell (2012) Future Disaster Response Action Plan

(Early Recovery – Through Cash Transfers). Cabinet Division,

Islamabad, Government of Pakistan.

FAO (2012) An Independent Evaluation of FAO’s Response to the July

2010 Floods in Pakistan. Accessed 10 January 2012:

http://www.fao.org/docrep/meeting/025/mC985E.pdf.

227 Sheeba Farooq

Fisher, D. (2010) Legal Implementation of Human Rights Obligations to

Prevent Displacement Due to Natural Disasters. Studies in

Transnational Legal Policy, 41, 551– 590.

Gilbert, C. L. and D. Vines (2006) The World Bank: Structure and

Policies (Vol. 3). Cambridge University Press.

GOP (2012) Working Paper for Funding for Excess Caseload under

CDCP Phase-II. Cabinet Division, Islamabad, Government of

Pakistan.

GOP (2013) Economic Survey of Pakistan 2012-2013. Finance Division,

Economic Advisor’s Wing, Islamabad, Pakistan.

GOP (2015) Annual Flood Report. Ministry of Water and Power,

Government of Pakistan: Islamabad.

Haider, M. (2013) World Bank Asks Pakistan to Improve Disaster

Management. The NEWS. 8 August, Page. No15.

Hanlon, J. (2004) it is Possible to Just Give Money to the Poor.

Development and Change, 35; 2, pp. 375-383.

House, S. (2012) Evaluation of Pakistan Flood Response 2011/2012:

Using Oxfam GB’s Global Humanitarian Indicator Tool. Oxfam.

Hunt, S. O’Leary, S. Newton-Lewis, T. and Z. Ali (2011) Evaluating

Implementation of Pakistan’s Citizen’s Damage Compensation

Programme (Phase I). Oxford Policy Management Limited,

Oxford: U.K.

IFAD (2013) The Sustainable Livelihoods Approach. Accessed 4 March

2013. http://www.ifad.org/sla/

IOM (2012) Do you hear me? Understanding Information Needs for

Disaster, Preparedness and Compensation. Islamabad: Pakistan.

Johansson, R. (2003) Case Study Methodology. In Proceedings of the

International Conference Methodologies in Housing Research,

Stockholm, 22-24 September.

Khan, A. N. (2013) Mountain Ranges in Pakistan. Pakistan Geographic.

Accessed 3 February 2013. http://www.pakistangeographic.com/

mountains.html.

Krantz, L. (2001) The Sustainable Livelihood Approach to Poverty

Reduction. An Introduction. SIDA.

Kundzewicz, Z. W., S. Budhakooncharoen, A. Bronstert, H. Hoff, D.

Lettenmaier, L. Menzel and R. Schulze (2002) Coping with

http://www.ifad.org/sla/

http://www.pakistan/


variability and change: floods and droughts. In Natural Resources

Forum. 26:4, 263-274.

Masud, N. and B. Yontcheva (2005) Does Foreign Aid Reduce Poverty?

Empirical Evidence from Nongovernmental and Bilateral Aid.

International Monetary Fund, 5-100.

GOP (2013) Pakistan Economic Survey 2012 2013. Accessed 21 March

2013.http://www.finance.gov.pk/survey/chapters_13/12- Populat

tion.pdf

Mirza, M. M. Q. (2003) Climate Change and Extreme Weather Events:

Can Developing Countries Adapt? In Climate Policy. 3:3, 233-

248.

Mosley, Paul, J. and Hudsonand S. Horrell (1992) “Aid, The Public

Sector and the Market in Less Developed Countries: A Return to

the Scene to Crime.” Journal of International Development, 4:2,

139-150.

Mwape, P. Y. (2009) An Impact of Floods on the Socio-Economic

Livelihoods of the People: A Case Study of Sikaunzwe Community

in Kazungula District of Zambia. Study (M.A). University of the

Free State, Africa.

NADRA (2013) Profile on Nowshera District. National Disaster

Management Authority, Prime Minister’s Secretariat, Islamabad:

Pakistan.

NADRA (2012) Mid-term review of CDCP. National Disaster


Pakistan.

NADRA (2012) Brief on Nowshera District. National Disaster


Pakistan.

NDMA (2012) Disaster Risk Management Needs Report 2012. National

Disaster Management Authority, Prime Minister’s Secretariat,

Islamabad: Pakistan.

OECD (2014) Engaging with the Public. 12 Lessons from DAC Peer

Reviews and the Network of DAC Development Communicators.

OECD.

229 Sheeba Farooq

Oxfam (2012) Restoring Livelihoods after Floods. Gender-Sensitive

Response and Community-Owned Reocovery in Pakistan.

Islamabad: Pakistan.

Oxfam (2013) Disaster Risk Reduction in Livelihoods and Food Security

Programming: A Learning Companion. Oxford: United

Kingdom.

Oxford Policy Management (2012) CDCP Mid-term review: Preliminary

Findings from Baseline Evaluation Survey. Oxford Policy

Management limited. Oxford: United Kingdom.

Oxley, M. C. (2013) A “People-Centred Principles-Based” post-Hyogo

Framework to Strengthen the Resilience of Nations and

Communities. International Journal of Disaster Risk Reduction,

4, 1-9.

Pakistan Budget (2012) Pakistan Economic Survey, 2011-2012, Accessed

10 March 2012:

http://cdn.onepakistan.com/content/2012/budget/12=PopulationL

abourForceAndEmployment.pdf.

Patton, M. Q. (2002) Qualitative Research and Evaluation Methods (3rd

ed.). Thousand Oaks.

Platteau, J. P. (2004) Monitoring elite capture in Community‐Driven

development. Development and Change, 35:2, 223-246.

Polastro, R., A. Nagrah, N. Steen and F. Zafar (2011) Inter-agency Real

Time Evaluation of the Humanitarian Response to Pakistan’s

2010 flood crisis. Madrid: DARA.

Régnier, P., B. Neri, S. Scuteri and S. Miniati (2008) From Emergency

Relief To Livelihood Recovery: Lessons Learned from Post-

Tsunami Experiences in Indonesia and India. Disaster Prevention

and Management, 17:3, 410-430.

Scoones, I. (1998) Sustainable rural livelihoods: a framework for

analysis), 22. Brighton: Institute of development studies.

Seck, P. (2007) Links between Natural Disasters, Humanitarian

Assistance and Disaster Risk Reduction: A Critical Perspective.

Occasional paper, Human Development Report, 2008.

Sekaran, U. and R. Bougie (2010) Research Methods for Business: A

Skill-Building Approach. John Wiley and Sons.

http://cdn.onepakistan.com/content/2012/budget/12=PopulationLabourForceAndEmployment.pdf

http://cdn.onepakistan.com/content/2012/budget/12=PopulationLabourForceAndEmployment.pdf


Semple, M. (2011) People’s experiences of the Pakistan floods and their

aftermath, July 2010 – July 2011. Friedrich Ebert Stiftung and

PATTAN. Islamabad: Pakistan.

Solidar. (2013) Pakistan Floods 2010. Islamabad: Pakistan. Studari.

(2004) Action Report toward Flood Disasters Reduction

Indonesian Case. International Flood Network.

Tahir, A. S., A.M. Malik and Z.A. Bhatti (2011) Post Flood Water

Quality Assessment Report (Sindh and Balochistan 2010).

Pakistan Council of Research in Water Resources, Islamabad.

(148-2011).

The NEWS (2013) Over 60 killed in heavy rains, flashfloods across

country. The NEWS. Editorial page. 4 August, p.1.

The NEWS (2013) The floods. The News, Editorial. 6 August, p.7. Editor:

Shaheen Sehbai. Islamabad, Pakistan.

UNDP (2010) Mainstreaming Disaster Risk Reduction into Development

at the National Level, A Practical Framework, UNDP, New York.

UNDP (2013) Early Recovery Needs Assessment. Islamabad: Pakistan.

UNOCHA (2010) Pakistan Floods Relief and Early Recovery Response

Plan. NewYork: United Nations.

WDR (2013) World Development Indicators (WDI) 2013. Washington

D.C.: The World Bank.

Wong, S. (2010) Elite Capture or Capture Elites? Lessons from

the'counter-elite'and'co-opt-elite'approaches in Bangladesh and

Ghana (No. 2010, 82). Working paper/World Institute for

Development Economics Research.

World Bank (2010) Responding to Floods in West Africa: Lessons from

Evaluation. World Bank: Washington D.C.

World Bank (2010) Pakistan Floods 2010: Preliminary Damage and

Needs Assessment. Asian Development Bank, World Bank and

Government of Pakistan. Islamabad: Pakistan, (58290).

World Bank (2011) Emergency Project Paper on a Proposed Credit in

the Amount of SDR 79.5 Million to the Islamic Republic of

Pakistan for a Flood Emergency Cash Transfer Project.

Islamabad: World Bank (59311-PK).

231 Sheeba Farooq

World Bank (2012) Pakistan Citizen’s Damage Compensation Program

(CDCP): Joint Supervision and Evaluation Mission (JSEM).

Aide-Memoire, July 9-13, 2012.

World Bank, Asian Development Bank and Italian Development

Cooperation. (2013) Citizen Damage Compensation Program.

Documentary. MCOM Pvt. Ltd, Islamabad: Pakistan.



Trade Costs of Pakistan with its Major Trading Partners:

Measurement and its Determinants

Saba Altaf*, Zafar Mahmood† and Shabana Noureen§

Abstract:

Trade costs are cited as an important determinant of volume of trade. Higher

trade cost is an obstacle to trade as it impedes the realization of gains from trade

liberalisation. Determinants of trade costs of Pakistan for the period 2003-2012 with

their major trading partners across Asia, European Union and North America are

investigated. Several gravity type variables have been used as trade cost determinants.

Trade costs for agricultural and non-agricultural sector are also calculated using a

micro-founded trade costs measure. Estimates of trade costs equivalents show a

declining trend of trade costs estimates over the period of study. Fully Modified

Ordinary Least Square estimation of the model shows that tariff rates and distances

between the trading partners increase the bilateral trade costs and thus adversely affect

trade. Results show that improvements in port infrastructure and membership of free

trade agreement significantly reduce the trade costs. Z-test shows that the effect of

determinants of trade costs for agricultural and non-agricultural sectors is invariant.

This paper recommends that the agreement on trade facilitation be implemented and

reduce the red tape at border crossings to cut down the trade costs.

Keywords: Trade Costs, Gravity Model, Liner Shipping Connectivity Index

1. INTRODUCTION

International trade is significantly affected by the trade costs

incurred locally and across the borders. Trade costs form a potentially

important barrier to trade. Higher trade costs are an obstacle to trade and

impede the realization of gains from trade liberalization,1 therefore

* Saba Altaf <[email protected]> is a graduate of Economics at School of Social

Sciences and Humanities (S3H), National University of Sciences and Technology,

Islamabad, Pakistan. † Zafar Mahmood <[email protected]> is Professor of Economics at School of

Social Sciences and Humanities (S3H), National University of Sciences and

Technology, Islamabad, Pakistan. §

Shabana Noureen <[email protected]> is a PhD Scholar at

Economics Department, School of Social Sciences and Humanities (S3H),

National University of Sciences and Technology, Islamabad, Pakistan.




233 Altaf, Mahmood and Noreen

special attention is given to trade costs. Owing to the importance of trade

costs in explaining the volume and direction of trade, international trade

economists are increasingly focusing upon trade costs and this has

become an area of key interest within the modern stream of international

trade research. The gradual decreeing of trade cost has been resulting

major risen in international trade thus this tremendous change has

brought improvement in every country for international trading over a

past years.

The pertinent question is what exactly are the trade costs? They

include all the costs incurred in getting a good to the final user, excluding

the marginal cost of producing the good itself. Hence, trade costs include

transportation costs (both freight costs and time costs), policy barriers

(tariffs and non-tariff barriers), information costs, contract enforcement

costs, costs associated with the use of different currencies, local

distribution costs (wholesale and retail) and legal and regulatory costs

[Singh, et al. (2014)].

Sources of trade costs are mainly divided into two main

categories. First category totally consists of bilateral factors to put

segregation between imported and exported and such factors are wide-

ly depending upon exogenous factors such as geographical distance,

common border or sharing a common language than particular policy

choices. The second category is composed of endogenous trade costs,

which are international connectivity such as air or maritime transport

services, tariffs and non-tariff measures, and other factors that facilitate

trade.

Evidence shows that with growing regionalism in the world,

countries have considerably reduced the tariff rates, i.e., 5% in

developed countries and 10 to 20% in underdeveloped countries

[Anderson and Van Wincoop (2004)]. With a drastic fall in tariffs on the

one hand, there are, on the other hand, some other barriers to trade that

are hampering the trade performance. Most important among those are

barriers relating to infrastructure quality besides the tariff and non-tariff

barriers, collectively these are referred to as policy barriers. Poor

1A growing literature has documented the impact of trade costs on the volume of trade

(see, for example, Anderson and Van Wincoop, 2004).

Trade Costs of Pakistan 234

institutions and poor infrastructure distort strategic trade policy focus,

not only in terms of the traditional mechanisms of tariffs and quotas but

also of infrastructure and logistics,2 the so-called ‘behind the border

issues’. Thus, besides the differences in economic size and endowments,

the differences in trade costs, which act as a friction to trade, is important

reason as to why some countries trade more than others.

In an increasingly globalized and networked world, trade costs

are of great importance from a policy perspective. This is because they

act as a determinant of the pattern of bilateral trade and investment as

well as of the geographical distribution of production. International trade

costs are large and vary widely across countries and sectors. These costs

are likely to be higher in developing countries as compared to the

developed countries due to the existence of substantial tariffs and non-

tariff measures accompanied by poor infrastructure, dysfunctional

transport and logistics.

Pakistan is a country heavily enriched with natural resources.

Pakistan’s major trade partners are Asian, European Union and North

American countries. These include China, USA, UK, India, Bangladesh,

Saudi Arabia, Malaysia, Japan, Germany and UAE. EU has now

emerged as Pakistan’s largest trading partner.3 Total trade between the

two amounts to about $10 billion with Pakistan’s share in EU market of

about 0.09% and the share of EU in Pakistani market is 11.39%. Pakistan

also has very strong trade ties with Asian economies like China, UAE,

Saudi Arabia, and Malaysia. The main reason behind massive trade of

Pakistan with Asian countries is low transportation costs, similarities of

consumer tastes and trading priorities. USA is also one of the strongest

trade partners of Pakistan.

The size of Pakistan’s current trade doesn’t truly reflect its trade

potential. This is mainly because the direction of Pakistan’s foreign

trade, which is trade cost dependent, has not changed virtually since its

independence. Keeping in view the trade potential of Pakistan and to

reap full benefits from international trade, it is thus imperative to have a

2See, for example, Khan and Weiss (2006), who explain how and why infrastructure

can assist the regional cooperation process. 3EUROSTAT (2013).


detailed insight into the determinants of trade costs. Pakistan needs to

pay serious attention to the trade costs because only then it will be in a

position to improve its ability to position better in global networks of

trade and production. A detailed study on the determinants and

calculation of trade costs will help identify the areas which need to be

given special attention to identify policies and measures that have a

significant effect on trade costs, and to prioritize them thus affecting the

overall trade flows and composition of trade consequently.

The research problem which is to be addressed and assessed in

this paper is “What are the factors that affect trade costs incurred by

Pakistan with its major trading partners”? The study uses a set of selected

trading partners of Pakistan due to the paucity of available data. The

main objective of the study is to measure the trade costs incurred by

Pakistan in agricultural and non-agricultural sector with its major trading

partners in three different regions of the world, i.e., Asia, Europe and

North America including USA, Germany, UK, Japan, China, UAE,

Saudi Arabia, Bangladesh, India and Malaysia and empirically

investigate the determinants of trade costs.

This area is virtually untapped in case of Pakistan. Therefore,

there is a need to have a research study that can show Pakistan’s position

in terms of trade costs and identify its determinants. Such a study can

provide insights that if properly targeted, trade costs can not only be

reduced but also proper policies can be formulated to help boost the

overall trade as well improving Pakistan’s position in global trade

network. This study would add to the literature by disaggregating trade

into two macro-sectors; agricultural and non-agricultural. Harmonized

System (HS)4 based on two digit level with its major trading partners in

three different regions of the world, i.e., European Union, Asia and

North America. The countries include USA, Germany, UK, Japan,

China, UAE, Saudi Arabia, Bangladesh, India and Malaysia.

4It is a coding system known for coding Harmonized Commodity Description of tariff

nomenclature. It is a system of International standard of names and codes in order to

classify traded products maintained by the World Customs Organization (WCO).


2. LITERATURE REVIEW

Trade costs have become a key area of interest for researchers.

In this regard, it is important to understand which factors trigger the trade

costs of a particular economy. Existing literature draws attention to some

of important determinants of trade costs. This section sheds light on the

existing literature in this area.

Limao and Venables (1999) identified the determinants of

transport costs and showed how they depend on geography and

infrastructure. Tobit model was estimated for the year 1990 taking 93

countries. Distance, contiguity and landlocked-ness were taken as

geographical determinants and quality of transport and communication

infrastructure were studied as infrastructural determinants. They

discussed that land distance is much more costly than sea distance.

Landlocked countries have high transport costs which can be reduced by

better infrastructure facilities. They further argued that trade volume can

be increased by a factor of five if transport costs are halved. The study

highlighted the cost of being landlocked as far as bilateral trade flows

are concerned.

Arvis, et al. (2007) estimated the cost attached to landlocked-

ness with regards to the international trade. Based on empirical analysis,

the study found out that large proportion of least developed countries are

landlocked and their market access depends upon the availability of

trade corridor or a transit system. High degree of unpredictability

associated with transportation time increases the trade costs of

landlocked economies along with high freight charges. The study

highlighted the need for reliable logistic services which are hampered by

flaws in implementation of transit system. They pointed that the business

community should design and implement comprehensive trade facili-

tation strategies. In addition to the physical constraints, least developed

countries are also faced with a problem of widespread rent seeking

activities. Thus, they showed that high trade costs of LDCs are mainly

due to high transportation costs which explain major proportion of high

logistic costs and vulnerability of supply chains and these areas need to

be targeted specifically.

Novy (2007) analysed the patterns of trade costs of UK and USA

with 31 trading partners from a period of 1960-2002. His study found


out that tariff equivalents of trade costs for USA have declined over the

period of study with US showing lowest trade costs with Canada and

Mexico while UK exhibited a remarkable increase in its bilateral trade

costs over time. Novy used micro founded trade cost measure for the

calculation of trade costs. The main determinants of trade costs were

classified into geographical, historical and institutional factors. Distan-

ce, landlocked-ness and exchange rate volatility ad tariffs showed a

positive relationship with trade costs while common border,

membership of free trade agreement negatively affected the trade costs.

Olper and Valentina (2007) examined the patterns of inter-

national trade costs in processed foods industry for a large cross section

of developed and developing countries over the period of 1976-2000.

Panel data estimation technique with country and time fixed effects was

used in this study. Tariff equivalents of trade costs were taken as the

dependent variable and the independent variables were divided into four

categories as geographical factors, historical and cultural linkage,

institutional factors and infrastructure development. Their study found

out that geographical and historical factors dominate the infrastructural

and institutional determinants of trade costs. Empirical results showed

that tariff equivalents of trade costs for the Emerging countries declined

by 13% over the period of study. However, developing countries showed

a low reduction pattern thus highlighting a need for government to focus

on the issue in order to achieve the goal of economic growth. They also

highlighted the need for freer trade environment keeping in view the

influential role played by trade policy in reduction of trade costs.

Duval and Utoktham (2011) examined the trade costs of Indian

Mekong sub-region and also evaluated the policy related and other

factors in order to facilitate trade and reduce trade costs. Novy’s trade

cost measure has been employed for calculating the trade costs. Various

trade related factors which possibly effected the trade costs of the Indian

Mekong sub-region were found to be bilateral distance between the

trading partners, cultural distance, tariffs between the trading countries,

liner shipping connectivity index, internet users per hundred people, ease

of doing business and monetary costs of moving a container from factory

to port and port to warehouse. Cross sectional data set of 64 countries


has been estimated for the year 2006 using the Ordinary Least Square

estimation.

The results show that trade costs between India and Mekong

countries are high. However, China, India, Thailand and most of other

Mekong countries are making progress in reducing trade costs among

themselves as compared to other countries like Japan and USA which is

mainly due to the enhanced regional connectivity. The study also

investigated the contribution of explanatory variables. Results revealed

that the natural barriers contribute about 22 percent to the total variations

in trade costs followed by the differences in maritime logistics and then

the trade related but non-trade specific measures such as credit

information, extent of information disclosure accounting for about 16

percent and 7 percent variations respectively in trade costs. The study

highlighted the importance of logistics and information technology

services regulation as important issues to reduce the trade costs.

Several researches have been conducted in different countries of

the world as far as measurement and determinants of trade costs are

concerned, but there is hardly any research on measurement of trade

costs of Pakistan and investigation of determinants of trade costs with

its major trading partners. Thus, the study at hand becomes all more

important to fill this research gap.

3. OVERVIEW OF THE ECONOMY IN THE CONTEXT

OF TRADE COSTS

Economic journey of Pakistan has faced serious global and

internal challenges since independence. Despite the critical circum-

stances, the country, however, managed to gain a momentum. In this

regard, the period of the 1960s was marked as the golden economic era

of Pakistan. Trade policies in that era focused on industrial development

and import substitution. Various incentives like tax rebates and exempt-

ions as well as export bonuses were offered on exports, which resulted

in a remarkable increase in export volume, with exports showing a

growth rate of 16.19%. In the late 1980s, due to increased economic

pressures and globalization forces, Pakistan initiated the process of trade

reforms and its intensity increased in the first half of the 1990s. Wide


ranging thorough liberalization programmes started in 1996-97 in the

agricultural sector. Government reduced average tariffs to a level of 15

percent compared to a high tariff rate of 51 percent in 1994/95 [WTO

(2007)]. Trade volumes of Pakistan increased sharply in the 2000s. Total

trade volume increased from $23,380 million in 2003 to $69,410 million

in 2013 [GOP (2014)].

Analysis of the trade costs of Pakistan for agricultural and non-

agricultural sector with its trading partners shows that on average

Pakistan is facing high levels of trade costs despite substantial fall in

tariffs worldwide. Quality of institutions and infrastructure differs across

countries thus causing a difference in their levels of trade and trade costs.

Therefore, today’s trade strategy goes beyond the traditional mechan-

isms of tariffs and quotas and includes “behind-the-border” issues, such

as the role of infrastructure and governance in supporting a well-

functioning trading economy. For instance, many studies show that

liberalisation of international transport services foster international trade

similar to tariff liberalization [Baier and Bergstrand (2001)].

Estimates of trade costs equivalents show that trade costs have

declined over the period of study thus showing an increase in

international trade volumes of the country (Table 1). It may be noted that

the agricultural sectors trade costs are comparatively higher than the non-

agricultural sector due to the existence of policy barriers including high

tariffs and non-tariff barriers. In addition, arguably the processing and

storage costs of agricultural commodities are higher than such costs on

industrial consumer goods.

Trade costs (TC) of Pakistan in agricultural and non-agricultural

sectors on average show a declining trend for the period 2003-2012

(Table 1). The reduction in trade costs (TC) is consistent with the

lowering of tariff rates. Tariffs not only make imports costly but also

discourage exports by raising the cost of imported inputs and act as an

implicit tax on exports. Thus, a fall in simple average tariff from 16.8%

in 2003 to 13.9% in 2012 has resulted into a rise in exports and imports,

also consistent with trade costs (TC) reduction.


Table 1. Trend in Trade Costs of Pakistan for Agricultural &

Non-Agricultural Sectors

Year TC Agr TC NAgr LSCI Pak Tariff ΔER

(Dep/App)

TV

(US$ million)

2003 204.08 159.93 19.29 16.81 0.008759 23380

2004 202.70 156.51 20.18 16.17 0.021759 27905

2005 197.26 154.50 21.49 14.61 0.012992 34989

2006 196.02 150.13 21.82 14.79 0.00785 45032

2007 198.41 150.55 24.77 14.9 0.160433 47516

2008 192.14 150.01 24.61 14.08 0.186123 59018

2009 193.70 151.84 26.58 14.78 0.049439 52510

2010 189.23 148.15 29.48 14.51 0.014068 54000

2011 190.39 147.83 30.54 14.25 0.082774 65224

2012 187.70 144.94 31.97 13.99 0.09536 68540

Source: Authors’ estimations, except for LSCI, average tariff based on World Bank (2013) and trade volume

based on GOP (2013).

Note: Positive change in exchange rate represents depreciation and negative change in exchange rate represents

appreciation. LSCI stands for liner shipping connectivity index represents Port infrastructure, TV represents the

trade volume.

An analysis of changes in the nominal exchange rate (ER) shows

depreciation of nominal exchange rate (ER) over the period of study.

Depreciation of exchange rate (ER) has increased the bilateral trade

flows relative to domestic trade, thus, causing a reduction in overall trade

costs (TC). Hence, depreciation of nominal exchange rate (ER) is seen

as a factor helping in trade costs reduction.

Reduction in trade costs can also be attributed to improvement in

port infrastructure and shipment. Table 1 shows a significant improve-

ment in liner shipping connectivity index (LSCI) from 19% in 2003 to

32 % in 2012. More than 95% of total freight trade of Pakistan is sea

borne; an improved and efficient port infrastructure facilitates trade and

reduces trade costs. Keeping this in view, Ministry of Ports and Shipping

of Pakistan is focused to achieve the objective of modernization and

corporatization of ports introducing modern technology and data base in

line with the present day trends, reviving ship-owning in the private

sector by removing the impediments, and enhancing tonnage and

profitability of Pakistan National Shipping Corporation. Fulfilment of

these objectives will further enhance port efficiency, reduce the costs for


port users and enhance port management accountability; consequently

reducing trade costs.

3.1. Sectoral Trade Costs

In trade costs equivalent terms, Pakistan and UAE on average

have the lowest levels of trade costs in their bilateral trade, i.e., 146.5%

for agricultural sector and 104% for non-agricultural sector. Table 2 and

3 provide trade costs of agricultural and non-agricultural sectors. There

are many factors behind these lower trade costs between two partners;

these include geographical proximity, cultural linkage, no currency

restrictions from UAE, abundant energy supplies, and no corporate

taxation [Hamid and Hayat (2012)]. Trade costs between two countries

are expected to decline further with the decrease in oil prices, which will

reduce transportation costs.

Another interesting finding of trade costs analysis is that despite

being neighbouring countries, tariff equivalents of trade costs between

Pakistan and India are quite high, i.e., 218% for agricultural and 176%

for non-agricultural sector (Tables 2 and 3). Trade costs are not low

between these two countries owing to the economic, political and

military tensions. There is discriminatory stringent application of non-

tariff barriers by India, i.e., regulatory and safety requirements that

dampens Pakistani exports to India. Political uncertainty, strict pro-

cedures for licensing permits and visa hassles also act as barriers to trade,

thus increasing trade costs. India follows a restrictive trade regime

especially in case of agricultural goods which is depicted by the high

trade costs of agricultural sector. Similarly, for textile exports, India

observes a large number of non-tariff barriers including para-tariffs,

sanitary and photo sanitary (SPS) measures and pre-shipment inspection.

Some goods can only be imported through specified ports and road

routes between the two countries are only open for exports of limited

number of commodities. These bottlenecks on road and rail route and

weak and inadequate transportation links between the two countries

further increases the trade costs. Also, Pakistan maintained a “Positive

list” for the Indian imports until 2011, which only allowed the imports

of these 1,946 items from India. Later on the approach of “Negative list”


was adopted by Pakistan which prohibited the import of 1209 items to

be imported from India. Positive list had also hindered the free flow of

goods between the two partners thus aggravating the overall trade costs

[Saleem, et al. (2014)]. With the adoption of a negative list, almost 85%

of goods can be imported from India compared to level of 25%

previously.

Pakistan and China are leading trading partners and neighbouring

countries, sharing a common border. However, bilateral trade costs

between two countries remain high. The government of China promotes

domestic consumption through structural tax reduction policies and there

is a strong domestic demand in China. Although, bilateral trade flows

between two countries are very large, yet China’s customs procedures

still require harmonization. Besides, its tariff regimes have not changed

substantially, which is a possible reason behind high trade costs.

In addition, China maintains restrictions, licensing and prohibit-

ions on grounds of state security and morality, all these factors add to the

levels of trade costs. Bilateral costs of trade between two countries can

be reduced by upgrading the Karakoram Highway which is the shortest

overland route between the two countries. Also, the construction of an

economic corridor is foreseen as a great opportunity to reduce the

staggering amount of time and distance consequently reducing the trade

costs. Long shipping routes between the two countries add to the costs

of trade which can be lessened by the construction of a direct corridor

from Kashgar to Gwadar, which is estimated to cut down the existing

costs associated to long distance by one-third of the current levels

[Kayani, et al. (2013)].

USA is also among the top ten major trading partners of Pakistan.

Trade costs between the two countries are high owing to the long

distance as well as many other contributing factors. USA’s domestic

trade relative to international trade with Pakistan is very high as

compared to Pakistan. The reason behind high values of domestic trade

is that there is an excellent working relationship between US

manufacturers and other distributors that provides wholesale customers

with access to barge product wherever and whenever they need it. Also,

there is an ease of transport (ground versus air/sea) which makes

domestic trade more feasible. Trade costs between two countries are high


because of large distance, stressed relationship between the government,

licensing and quality control requirements from USA government.

Pakistan is a country that is included in the list of Restricted Entities by

USA, imposition of non-tariff barriers makes textile and clothing

products of Pakistan suffer the most.

Table 2. Estimates of Trade Costs Equivalents for Agricultural Sector US Dollar (USD)

Year IND UAE CHN SA UK USA MYS JPN GMY BD

2003 241.26 148.28 222.76 217.26 190.21 205.22 200.31 239.38 224.92 151.18

2004 239.87 148.19 219.47 215.38 194.16 186.24 201.76 242.01 224.27 156.65

2005 217.54 142.31 207.87 210.65 191.83 201.09 199.33 240.87 220.86 150.09

2006 196.67 144.37 217.21 208.85 188.42 192.01 203.25 244.09 216.73 158.92

2007 222.88 148.86 196.07 208.87 190.65 193.84 199.07 233.54 218.59 151.07

2008 218.38 147.27 207.21 197.99 185.16 181.39 192.41 233.81 214.41 151.24

2009 221.74 146.92 206.35 199.52 184.25 190.03 180.85 229.68 212.57 150.12

2010 219.91 142.73 203.41 196.37 182.96 188.95 185.80 226.03 208.57 149.04

2011 213.50 149.18 199.91 206.34 181.23 192.93 179.77 219.32 209.42 149.82

2012 212.88 147.20 191.78 202.98 180.01 187.63 179.84 214.91 208.68 149.01

Avg. 218.66 146.53 206.62 204.42 192.68 191.93 192.24 232.57 209.76 151.71

Source: Authors’ calculations.

Note: IND stands for India, CHN stands for China, BD stands for Bangladesh, SA stands for Saudi Arabia, MYS

stands for Malaysia, JPN stands for Japan, GMY stands for Germany.

Pakistan and EU enjoy very strong and rapidly growing trade ties.

We have chosen two countries from EU, United Kingdom and Germany

for the purpose of trade costs analysis. Estimated trade costs show that

despite the fact there is no cultural or geographical proximity between

Pakistan and selected EU member states, trade costs on average are not

very large. A further decline in trade costs is expected to occur by the

GSP plus status granted by EU to Pakistan in 2014. Before that, Pakistan

was given a general GSP status, and Pakistani exports faced some sort

of non-tariff barriers like standard intellectual property rights, rules of

origin and competition policy.

Pakistan and Bangladesh have not been able to bring about a

significant reduction in their bilateral trade costs. Though trade between

two countries is growing progressively and has crossed $1 billion mark

but there is a need to develop trade facilitation strategies that can further


reduce trade costs. At present there is no direct air link between two

countries, especially between Lahore and Dhaka. Infrequent shipping

arrangements between the two countries hamper flow of goods between

Pakistan and Bangladesh.

Japan is an important trading partner of Pakistan. There is a huge

potential for further increase in trade volume between the two countries.

Trade costs estimates, however, do not present a very encouraging

picture. Trade costs equivalents are very high. The import regulations,

applicable standards and quarantine requirements make it all the more

difficult to export Pakistani products specially food items. Pakistani

exports also have the disadvantage of being more distant from the market

than its competitors, such as China, Russia, Thailand, South Korea, etc.

This not only increases transportation costs but also delays the delivery

of goods, whereas Japanese importers prefer small size lots with short

delivery schedules. Both the countries need to overcome these

impediments to bilateral trade.

In the modern time, importance of trade costs as a determinant of

national trade performance and competitiveness has been seriously

Table 3. Estimates of Trade Costs Equivalents for

Non-Agricultural Sector US Dollar (USD)

Year IND UAE CHN SA UK USA MYS JPN GMY BD

2003 203.47 107.78 162.67 150.15 157.63 171.20 164.49 161.65 180.25 139.10

2004 185.64 109.32 160.49 147.19 160.33 165.90 160.30 160.76 176.01 139.18

2005 182.60 106.39 153.83 148.71 158.85 164.03 160.52 156.64 176.86 136.60

2006 162.07 105.77 148.30 144.82 152.83 165.49 161.97 151.43 175.02 133.68

2007 167.94 104.07 146.72 138.61 160.79 166.59 155.62 157.19 174.32 133.98

2008 164.32 103.23 153.62 137.68 157.70 165.90 151.17 156.63 175.14 134.78

2009 174.46 103.62 158.80 135.74 157.75 163.74 149.79 164.01 172.78 137.91

2010 171.49 103.06 152.93 130.70 147.88 163.32 145.96 159.42 171.89 133.82

2011 169.34 101.52 154.58 131.80 148.71 163.63 147.88 158.34 171.37 131.54

2012 172.11 100.48. 154.64 127.84 139.88 158.63 142.73 160.10 162.04 130.99

Avg. 176.12 103.72 154.66 139.33 154.20 164.01 154.04 158.61 173.57 134.24

Source: Authors’ calculations.

Note: IND stands for India, CHN stands for China, BD stands for Bangladesh, SA stands for Saudi Arabia, MYS

stands for Malaysia, JPN stands for Japan, GMY stands for Germany.


recognized by the developed countries. Their governments have been

critically analysing and performing research for making effective

policies for reduction of trade costs. On the other hand, developing

countries have been rather ignorant and little efforts have been made so

far at policy level to address this issue. Pakistan is not different from

other developing nations. By looking at trade costs estimates, we find

that the country still faces high bilateral trade costs viz a viz its major

trading partners. This shows government’s lack of policy attention

towards trade facilitation. Pakistan still exports large volume of agri-

cultural products, while trade costs for agricultural sector are sub-

stantially higher than that of non-agricultural sector, which speaks of

sectoral inefficiency and bias in policies. Thus, the key need is to identify

the primary sources of trade costs and formulate what government

should do to address them so that trade can be used to sustain high rate

of economic growth over a longer period of time.

4. THE TRADE COSTS MODEL

4.1. Theoretical Framework

Trade costs are cited as important determinant of international

trade. Given the nature and pattern of trade costs, the Gravity model of

international trade is most suitable to determine factors that affect trade

costs. This is because the model provides main link between trade flows

and trade barriers. The Gravity model has become a major pillar in

applied international economics [Evenett and Hutchinson (2002)]. It is

basically motivated by the Newton’s gravitational law in which the

gravitational force utilized among two bodies is determined by their

distance and mass. This model became popular in international

economics with the pioneering work of Tinbergen (1962). It relates

bilateral trade flows to the GDP, distance, and other factors including

trade barriers. Anderson (1979), Deardoff (1998), Hummels (1999),

Baier and Bergstrand (2001), Limao and Venables (2001) have applied

it in a wider sense to infer trade flow effects of institutions such as

customs unions, exchange rate mechanisms, ethnic ties, linguistic

identity and international borders.


This paper makes use of Novy’s (2008) trade costs measure. This

is a micro founded measure of trade cost that has been derived from

Anderson and Van Wincoop (2003) model based on the Gravity

equation. The Gravity equation has been most widely used instrument

for modelling the bilateral trade flows. As a workhorse of international

trade, it relates countries bilateral trade with their economic sizes and

trade costs. This measure analytically solves the theoretical gravity

equation for the trade cost parameters that capture the barriers to

international trade.

Novy (2007) explained multilateral trade hindrance factors in

detail and solved that trade function too. These new strategies are

applicable to both international and domestic trade resistance. Basically,

when the cost of a particular product reduces then such items are shipped

out of countries and this implies that such hindrance have huge impact

on domestic trade too. Previous theories don’t justify this boarder

hindrance and also, they don't take domestic trade in any account. A

slight change in trade barriers can bring noticeable change in resources

and can shift recourses into tradeable and non-tradable sectors and this

will result in changes in trade flows (either bilaterally or multilaterally).

Hence, multilateral resistance of the trading countries explains domestic

trade very well so it is important to include domestic factor into the

equation also to address the home biased.

The motivation behind Novy’s approach was to overcome the

drawbacks that were associated with the theory-based gravity framework

by Anderson and Van Wincoop (2003), which imposed certain arbitrary

trade cost functions. The theory-based gravity formulation was a

refinement of the traditional gravity equation to include multilateral trade

resistance variables.

Anderson and Van Wincoop [AvW (2003)] derived a micro

founded trade cost measure based on a multi-country general equilibrium

model expressed as:

... (1)

where, 𝓧𝒊𝒋 is the bilateral trade from i to j, 𝓨𝓲 & 𝓨𝓳 are nominal income

of country i and j, 𝓨𝔀 is the world income, 𝚷𝓲 is the outward multilateral

resistance of country i, 𝓟𝓳 is the outward multilateral resistance of


country j, and 𝓽𝓲𝓳 is the bilateral trade cost measure, σ is the elasticity

of substitution between goods. The main innovation in AvW's (2003)

model is to incorporate exporter and importer price indices (Π and P )

such that trade not only depends on bilateral trade costs between the two

countries but also on the trade “resistance” they face with all of their

trading partners in the rest of the world. That is, country i is more likely

to trade with country j if πi is higher, meaning the multilateral resistance

of country i to all other partners is higher.

Using Equ. (1), consider the intra-national trade of country i as:

… (2)

and rewrite it as:

… (3)

which solves for country i's multilateral resistance. Multiplying Equ. (1)

with 𝓧𝓳𝓲, we obtain:

… (4)

substitute Equ. (3) for country i and j into (2), we can derive the bilateral

trade costs relative to domestic trade costs expressed as tariff equivalent

by subtracting 1:

… (5)

where,

τij = tariff equivalent trade cost (i.e., measures domestic trade relative to

bilateral trade).


tij = international trade costs from country i to country j.

tji = international trade costs from country j to country i.

tii= intra-national trade costs of country i tjj denotes intra-national trade

costs of country j.

xij =international trade flows from country i to country j.

xji =international trade flows from country j to country i.

xii =intra-national trade of country i.

xjj =intra-national trade of country j.

σ denotes elasticity of substitution between all goods.5

τij is defined as a ratio of trade cost across national border relative to

trade cost within national border weighted by the elasticity of

substitution. It must be noted that τij is not directional, i.e., τij measures

the barrier between country i and j on average, so that it is a two-way

trade cost measure. Intuitively, it measures the bilateral trade cost for

both importing and exporting countries. Trade costs τij, thus represent the

geometric average of international trade costs between countries i and j

relative to domestic trade costs within each country. Intuitively, trade

costs are higher when countries tend to trade more with themselves than

they do with each other, i.e., as 𝑋𝑖𝑖𝑋𝑗𝑗/𝑋𝑖𝑗𝑋𝑗𝑖 increases. As the ratio falls

and countries trade more internationally than domestically, international

trade costs must be falling relative to domestic trade costs.

An additional advantage of Novy’s trade cost measure is that it

allows time-varying measurement of bilateral trade barriers. With

readily available trade and production data in tradable goods categories,

we are able to measure and explain the determinants of bilateral border

effects.

The gravity equation represents one of the greater successes in

empirical economics, as it describes the value of bilateral trade, which is

function of the market size of the importer as well as exporter, and

5See, Anderson and Van Wincoop (2003) for detailed discussion on elasticity of

substitution between goods. This thesis follows AVW’s and Novy (2008) σ=8, which

is the middle point of available estimates. Smaller value of σ results in higher trade

costs showing that consumers are irresponsive to prices and trade costs and consume

larger amounts of foreign goods.


distance among them [Lili (2011)]. Market sizes embody push and pull

factors that affect value of trade flows, and are usually characterized by

the GDP. Distance is generally measured by geographic distance among

two regions (absolute distance). It is anticipated that large distance

between trading partners leads to a decrease in trade, as trade will

become complicated and bring transaction costs. The basic Gravity

model is as the following:

𝑇𝑖𝑗 = 𝐺 (𝑌𝑖𝑌𝑗

𝐷𝑖𝑗) … (6)

where, Tij is bilateral trade volume, for sum of exports and imports; Yi

is country i's GDP; Yj is country j's GDP, Dij is the distance among

country i and country j; and G is a constant; and is independent of any

subscript as it links to a standard Gravity equation in the following form.

The multiplicative nature of Equ. (6) Suggests that by taking logs it can

be made linear in parameters:

lnTij = lnG + a1lnGDPi + a2lnGDPj - a3lnDij + εij … (7)

Objectives of this paper are to test the following two hypotheses:

H1: Connectivity constraints are more important trade deterrents than

tariff barriers.

H2: Determinants of trade costs have similar effect on agricultural and

non-agricultural sectors.

The relationship between trade costs and its determinants is

difficult to capture given the paucity of data on all the factors involved.

However, in order to explore the determinants of trade costs, our

empirical analysis has used several gravity-type variables including

distance, infrastructure development, exchange rate, tariff, area and two

dummy variables for contiguity and free trade agreement between the

trading partners.


4.2. Empirical Model

Following Novy (2007), joint observation of non-bilateral

variables for country i and j are constructed by multiplying the single

country variables to lead to symmetric and constant interaction effects.

All variables are taken in the log natural form.

τij = ƒ (DIST, TARIFF, EXCH, LSCI, AREA, CONT, FTA) … (8)

where, τij is the dependent variable representing tariff equivalent of

trade costs, DIST is the distance among Pakistan and partner country,

TARIFF is the product of tariffs imposed by Pakistan and other trading

partner, EXCH is the official exchange rate with respect to Pakistan

(taken in current US dollars), LSCI is the linear shipping connectivity

index of Pakistan and partner country, AREA represents product of land

area of two trading partners, CONT and FTA are dummies for

contiguity and free trade area, which take the value one if two partner

countries are contiguous and members of FTA and zero otherwise.

Distance appears in the Gravity model as proxy of remoteness or

transportation costs implying that coefficient of distance is expected to

have a positive impact on trade costs. This paper uses liner shipping

connectivity index (LSCI) as a measure of infrastructure development of

the trading countries. Our model includes a dummy variable to show

common border with the trading partner. Those countries that share a

common border are reflected by a unitary value of dummy variable,

known as contiguity. Common border again is a proxy for transportation

and information costs, which tend to be lower for contagious trading

partners as they are well aware of consumer’s choices and trading

prospects, thus making mutual trade less costly. Coefficient of conti-

guity is expected to be negative.

Ample land is an indicator of big economy and bigger population

with high domestic demands. In order to fulfil that high demand foreign

goods are also accepted and larger countries have cultural diversity,

residents have greater acceptability for a variety of cultures, which calls

for greater imports [Saleem and Mahmood (2014)]. Thus, trade

increases and overall trade costs decrease. Coefficient of area of trading


partners is expected to have a negative sign. Another dummy has also

been included to evaluate the effect of Free Trade Area (FTAs) on trade

costs. Dummy for FTA is expected to have a negative impact on trade

costs.

Tariffs and exchange rate are two policy related or institutional

determinants of trade costs. Tariffs imposed by partner countries are

used as a measure of restrictiveness to trade flows. Aggravation of tariffs

imposed by the trading partners is expected to increase the bilateral trade

costs, not only it affects imports but the level of exports also declines if

tariffs are imposed on raw materials. Issues of duty draw back further

add to the level of trade costs. Thus, overall international trade declines

and intra national trade increases consequently increasing trade costs.

Exchange rate is used as a measure of competiveness in

international trade flows. The study uses official exchange rate as a

determinant of trade costs. Increase in nominal exchange rate leads to an

increase in overall volume of trade is a well-established fact. An increase

in trade flows with nominal depreciation therefore leads to decline in

trade costs as trade flows and trade costs are inversely related. Keeping

this in view, coefficient of exchange rate is expected to have a negative

sign.

4.3. Empirical Specification

The general empirical model reported in Equ. (8) is transformed

as the following econometric equation, which links tariff equivalents of

trade costs with its determinants and is given as:

τij = β0 + B1EXCHijt + B2TRit*TRjt + B3DISTij + B4LSCIit*LSCIjt +

B5CONTij + B6AREAi*AREAj + B7FTAijt + εijt … (9)

In our opinion, model in Equ. (9) will help us determine the

impact of these variables on trade costs of Pakistan. The findings from

this model will have important implications for the policy, as it will help

the policy makers to figure out those areas that can bring about

significant reductions in trade costs and prioritize policies accordingly.


Table 4. Definition and Sources of Data Variables

Variable Definition Proxy of Data Source

Export/

Import

Bilateral trade flows between

country i and j

Direct Variable UN Comtrade

GDP Output of agricultural and non-

agricultural sectors of country i and

j in current US Dollars

Direct Variable WDI, World Bank

TARIFF Product of simple average tariffs

imposed by Pakistan and partner

country.

Measure of

restrictiveness

WDI, World Bank

EXCH Average official exchange rate of

Pakistan (US Dollar)

Competitiveness Pakistan Economic

Survey, GOP

DIST Distance between Pakistan and

partner countries capital cities.

Transportation

costs

CEPII

AREA Product of country i and j land area. Size of

economy

CEPII

FTA Dummy equal to unity if two

countries are a member of free trade

area.

Market access WTO website

CONT Dummy equal to unity if two

countries share a common border.

Information

costs

CEPII

LSCI Product of country i and j scores on

liner shipping connectivity index.

Trade

infrastructure

WDI, World Bank

5. RESULTS AND DISCUSSION

5.1. Summary Statistics

Summary statistics is a quantitative description of the main

features of the data used in the study. Mean and median are used as

measures of central tendency while standard deviation, maximum and

the minimum values represent measures of variability. Table 5 provides

summary statistics of Pakistan’s trade costs with reference to the

variables included in the study. A fleeting look at the summary statistics

shows that highest mean value of total trade costs for Pakistan is

138.50% with maximum value of 191.8% and minimum value of 98.4%.

To identify whether a long run relationship between trade cost

and explanatory variables exists or not, the prerequisite is to analyse the

time series properties of all the variables first. As the co-integration tests


Table 5. Summary Statistics

Variable Mean Median Std. Dev Minimum Maximum Observations

TC 138.50 135.74 22.702 98.4 191.8 140

TARIFF 0.924 0.633 0.737 0.205 4.59 140

DIST 4299.13 3916.826 2472.13 683.369 11091 140

AREA 1042615 625300.8 873038.8 464239 2907092 140

LSCI 36.064 37.697 13.466 9.504 68.146 140

Dummy Variables

CONT 0.214 0 0.410 0 1 140

FTA 0.285 0 0.451 0 1 140

can only be performed when the panels are non-stationary. For the

purpose of checking the stationarity of the series, panel unit root test

[Levin, Lin and Chu (2002)] is run on the basis of the following

hypothesis:

H0: Variables exhibit a unit root.

H1: All the variables are stationary.

5.2. Empirical Results

5.2.1 Empirical Results of Pooled Unit Root Test

In order to check the presence of unit root in selected countries,

pooled unit root test is conducted using the Eviews-8 Software. Table 6

reports the results of Levin, Lin and Chu (2002) stationarity test showing

that the variables TC, TARIFF, EXCH and LSCI are stationary at the first

difference, i.e., I(1). Distance between countries and country area fail to

show any result because they are independent of time. Remaining two

variables included in the model are dummy variables.

5.2.2. Empirical Results of Kao (1999) Co-integration Test

To determine whether variables with first difference orders of

integration, i.e., I(1) yield spurious regression or a long run relationship

does exist, Kao (1999) panel co-integration test is run based on the null

hypothesis of no co-integration. Table 7 shows that the null hypothesis


of no co-integration is rejected thus confirming that a long run relation-

ship does exist. In other words, the possibility of spurious results has

been ruled out.

Table 6. Levin, Lin & Chu Test for Stationarity

Variable

Level First Difference

Order of Integration

Stat. Prob. Stat. Prob.

TC -1.315 0.0946 -8.717 0.0000 I(1)

TARIFF 0.151 0.5603 -4.736 0.0000 I(1)

EXCH -28.001 0.0000 I(1)

LSCI 0.370 0.6444 -5.273 0.0000 I(1)

Table 7. Kao (1999) Residual based Co-integration Test Estimation

Kao Co-integration Test Dependant variable: D (RESID)

Included Observation: 139 after

adjustment

Variable Coefficient t-statistic P-value

RESID(-1) -0.233 -3.749 0.0003

Null Hypothesis: No co-integration

The results of Kao (1999) co-integration tests confirm the

existence of a long run relationship between the dependent and

explanatory variables. Therefore, the application of OLS technique will

yield biased and inconsistent estimators. Fixed effects model cannot be

applied to models involving time invariant variables such as distance as

it leads to problem of endogeneity. We thus need to adopt an alternative

method to estimate the co-integrated panel. In this regard, Panel Fully

Modified Ordinary Least Square (FMOLS) method was developed by

Pedroni (1996), which uses a correction approach to deal with the

nuisance parameters and thus gives long run coefficients for the

estimated model correcting for endogeneity and serial correlation.

FMOLS has an advantage over other techniques as it allows for

estimation of common co-integration vectors while allowing for

heterogeneity both across time and cross sections [Pedroni (2004)].


Thus, the resultant estimates are more consistent, free of serial

correlation and endogeneity.

5.2.3. Empirical Results of FMOLS: Total Merchandise Trade

Results of fully modified ordinary least square model show that

trade costs equivalents for selected trading partners of Pakistan are

significantly dependent on the explanatory variables included in the

model.

Table 8 shows the estimated results of determinants of trade costs

for overall merchandise trade with Pakistan’s major trading partners.

Dependent variable is the log of trade costs equivalents for total

merchandise trade.

The results depict that nominal exchange rate (EXCH) is

statistically significant at 1% level and has a negative sign. There is an

inverse relationship between depreciation of nominal exchange rate and

trade costs. In other words, with depreciation of the exchange rate, total

volume of trade rises. As trade goes up, intra-national trade goes down

resulting into a decline in trade costs. The coefficient for exchange rate

suggests that 1% depreciation of exchange rate reduces trade costs by

0.03% (Table 8). This result is consistent with the findings of Singh, et

al. (2012). Thus, an increase in international trade greater than the

increase in intra-national trade as a result of currency depreciation

implies that it has become easier for countries to have more trade

internationally rather than trading internally, which is tantamount to a

decline in trade costs. It is pertinent to note that with depreciation of

nominal exchange rate it is realized that the growth in total bilateral trade

with selected countries over the period of 2003-2012 is 50.4%, which is

larger than the growth of intra-national trade which increased by 37.2%.

Tariffs always act as an obstacle to international trade, thus,

increasing the trade costs. Imposition of tariffs not only decreases the

level of imports as well as exports, because tariffs imposed on imported

raw materials and inputs used in production of export tables, causing a

switch towards intra-national trade leading to increase in trade costs.

Thus, increase in tariffs adversely affects overall trade flow. Here,

product of tariffs imposed by Pakistan and its trading partner is used,


reflecting degree of market access in two countries, which leads to

increase in trade costs. Estimated coefficient sign for this variable is

positive but is statistically insignificant. Results show that 1 % increase

in tariffs will increase the trade costs by 0.02% (Table 8). These results

are in line with the findings of Novy (2013) and Wincoop, et al. (2004).

* Significant at 1%, ** significant at 5%.

Distance, area and common border are geographic determinants

of trade costs. Distance between the trading partners affects the physical

transport cost. Geographic distance between the trading countries is a

hindrance to bilateral trade flows. Estimated results show that

geographic distance between Pakistan and its trading partners is

positively related to the trade costs (Table 8). It indicates that 1%

increase in distance increases the trade costs by 0.28%. Our result is

consistent with the study of Duan and Jason (2012).

The regression coefficient of the variable land area of Pakistan

and the trading partner is negative and statistically significant at 1%

Table 8. Empirical Results of FMOLS: Total Merchandise Trade

Variable Coefficient Std. Error t-Statistics p-values

TARIFF( TRi*TRj) 0.019663 0.051249 0.383675 0.7020

EXCH -0.028292 0.009666 -2.927088 0.0042*

LSCI( LSCIi*LSCIj) -0.179337 0.045527 -3.939140 0.0002 *

DIST 0.278061 0.071994 3.862296 0.0002 *

AREA(Areai*Areaj) -0.047967 0.015814 -3.159048 0.0031 *

CONT -0.131884 0.099674 -1.323153 0.1888

FTA -0.166789 0.071971 -2.31745 0.0224**

C 2.644451 0.553275 4.779636 0.0000*

R-squared 0.678589 Mean dependent var 4.916574

Adjusted R-squared 0.656313 S.D. dependent var 0.181657

S.E. of regression 0.106496 Sum squared resid 1.145483

Durbin-Watson stat 0.592009 Long-run variance 0.028958


confidence level. It implies that when there is 1% increase in land area,

trade costs decrease by 0.04% (Table 8). Intuition behind this result is

that generally countries with large land areas have large economies and

populations, and thus have high domestic demand. To fulfil domestic

demand, foreign goods are also accepted by local population, which

results into trade. Moreover, in large size countries, cultural diversity is

a hallmark and the residents have greater acceptability for a variety of

culture including foreign cultural goods, which also causes greater

import of cultural goods [Salim and Mahmood (2014)]. Thus, inter-

national trade flow increases and trade costs decrease. The present

study’s empirical result is same as that of Lili (2011).

Liner shipping connectivity index (LSCI) is used as a proxy for

infrastructure development. Estimates of regression show that LSCI has

a negative and statistically significant impact on trade costs. Better

maritime connectivity and port efficiency reduce the level of delays in

shipment of goods and thus lower trade costs. Results show that 1%

increase in LSCI decreases the trade costs by 0.17% (Table 8). These

results corroborate with the findings of Duval and Chorthip (2010),

Singh, et al. (2012) and Olper and Valentina (2007).

Dummy variable for free trade agreement (FTA) exhibits a

negative and significant relationship with trade costs. According to the

regression results, Pakistan’s membership in a free trade area reduces

trade costs by 0.16% (Table 8). Free trade area reduces barriers to

exchange and increases international trade through specialization, divis-

ion of labour and comparative advantage. Thus, an increase in

international trade in the aftermath of free trade agreement reduces trade

costs. Our results are in line with the findings of Novy (2007).

R-square is used to measure the regression’s success in

determining the values of dependent variables. Overall, our model per-

forms reasonably well and about more than half of the variation in

dependent variable is being explained by independent variables.

Adjusted R-square is 0.66, which shows that the above determinants are

explaining 67% of variation in trade costs. Standard deviation of

dependent variable is less than which indicates greater reliability of the

results. Model is also adjusted for serial correlation and possible

endogeneity problem because of FMOLS.


Estimation results for trade costs equivalents for agricultural and

non-agricultural sectors along with the z-test results of cross model

coefficients comparison are given in Appendix, the included explanatory

variables yield same statistical relationship with the dependent variable

as in the case of total merchandise trade.

6. CONCLUSION AND POLICY IMPLICATIONS

This study analysed the estimates of trade costs for overall trade,

agricultural trade and non-agricultural trade of Pakistan with its major

trading partners across Asia, Europe and North America over the period

2003-2012. Moreover, it examined the relationship between trade costs

and its major determinants using the panel data estimation techniques.

The study adds to the literature by disaggregating trade into two macro-

sectors, agriculture and non-agriculture. Existing studies only used total

trade, without attempting on sectoral trade details.

Despite the fact that international economy has considerably

integrated, our analysis of tariff equivalents of trade costs emphasises

that large unexploited gains can be reaped by further reducing the wedge

between the cost of producing a good and price paid by ultimate

consumer, i.e., by cutting down the trade costs.

Our estimates of trade costs reveal that Pakistan’s trade costs are

following a disproportionate pattern with its trading partners. Although,

the estimates show a considerable reduction in trade costs, yet they

indicate that substantial room remains for lowering them further. High

bilateral trade costs with some of its very largest trading partners in

particular calls for policies that can effectively reduce trade costs

between the trading partners. Policy makers need to address the

dynamics of higher trade costs in order to improve country’s absolute

and relative position in the global trade.

At the sectoral level, costs of trade for agricultural sector tend to

bypass the costs of trade for non-agricultural sector. The fact that

agricultural trade costs in many developing countries are relatively

larger than that of the non-agricultural sector suggests that focusing on

trade facilitation efforts for agricultural sector would be particularly


productive for Pakistan as WTO’s agreement on trade facilitation also

emphasis on the release of perishable goods at the earliest possible.

In addition to mapping out the level of trade costs of Pakistan in

the recent decade, we used econometric method to investigate various

determinants of trade costs. For this purpose, we decompose the trade

cost components into various policy and non-policy features. A key

finding is that distance, maritime transport and trade facilitation matter

for trade costs. Two areas which are highly amenable to policy

intervention for reduction of trade costs are the trade infrastructure and

free trade areas with the trading partners. UNCTAD’S liner shipping

connectivity index is a more important source of trade costs than tariffs.

This is because better shipment connectivity with the trading

partners efficiently improves transportation routes thus reducing time

and other costs. Similarly, we find that free trade agreements also play a

significant role in reducing the costs of trade; this implies that the FTAs

of modern era including a fall of non-tariff and behind the border

regulatory measures will be helpful to achieve the target of trade costs

reduction. Empirical analysis allowed to identify those trade facilitation

measures and policies which are most effective determinants of trade

costs. It suggests that an increase in geographical distance between

trading partners, and tariffs are positively linked with the trade costs.

However, land area and common border between trading partners,

nominal depreciation of exchange rate, liner shipping connectivity index

and membership of a free trade area all because a decline in trade costs.

The benefits of trade as an engine of economic growth and

sustainable development as well as means of poverty reduction can only

be achieved if these high trade costs are controlled. Higher trade costs

lower the competitiveness, thus limiting the potential benefits of trade.

Pakistan is a developing country and trade can turn out to be a helpful

instrument to achieve sustainability and economic welfare provided

these large trade costs are taken care of.

The study evidently shows that there is ample room for reduction

in trade costs if proper policy actions are taken. Findings of the study

have the following implications for policy making:


Pakistan should actively participate in WTO’s agreement on

trade facilitation and reduce the red tape at border crossings to

cut down the trade costs.

Shipment of perishable agricultural goods must be expedited and

releasing these goods at the earliest could help reduce trade costs.

Improve port connectivity, cargo handling and means of

transportation, i.e., roads, railways and air links.

In addition to tariff reduction, NTB’s must be streamlined and

harmonized to reduce trade costs.

Effect of longer distance can be limited by the development of

both hard and soft infrastructures by applying modern

technological methods: internet, publicity campaigns and

electronic media.

Initiation of mega projects like CPEC can bring about the much

needed trade costs reduction.


APPENDIX

A. Empirical Results of FMOLS: Agricultural Sector Trade


TARIFF( TRi*TRj) 0.072242 0.052622 1.372851 0.1732

EXCH -0.052983 0.011580 -4.575206 0.0000*

LSCI( LSCIi*LSCIj) -0.171619 0.054041 -3.175734 0.0020 *

DIST 0.149250 0.071237 2.095166 0.0389 **

AREA(Areai*Areaj) -0.059225 0.012518 -4.731075 0.0000 *

CONT -0.040529 0.030851 -1.313729 0.1922

FTA -0.144177 0.067401 -2.139099 0.0351**

C 4.578977 0.260851 17.55403 0.0000*





*significant at 1%, ** significant at 5%.

B. Empirical Results of FMOLS: Non-Agricultural Sector Trade


TARIFF( TRi*TRj) 0.004906 0.037630 0.130386 0.8965

EXCH -0.027262 0.013489 -2.021037 0.0462**

LSCI( LSCIi*LSCIj) -0.186992 0.070409 -2.653159 0.0094 *

DIST 0.282673 0.069443 4.070551 0.0001*

AREA(Areai*Areaj) -0.049181 0.012387 -3.970405 0.0001*

CONT -0.009774 0.079129 -0.123524 0.9020

FTA -0.292959 0.077992 -3.756287 0.0003*

C 6.164287 1.114648 5.530256 0.0000*





*significant at 1% ,** significant at 5% .


C. Z-scores for Cross Model Coefficients

Variable Calculated Z- score

Tariff 1.041

EXCH -1.44

LSCI 0.173

DIST 4.33

AREA 0.570

CONT 0.36

FTA 0.828

REFERENCES

Anderson, J.E. (1979) A Theoretical Foundation for the Gravity

Equation. The American Economic Review, 69: 106-116.

Anderson, J.E. and E.V. Wincoop (2003) Trade Costs. Journal of

Economic Literature, 42:3, 691-751.

Arvis, J.F., Y. Duval, B. Shepherd, and C. Utoktham (2013) “Trade

Costs in the Developing World: 1995-2010.” World Bank Policy

Research Working Paper, (6309). Available at:

http://www10.iadb.org/intal/intalcdi/PE/2013/.pdf12708.

Baier, S.L., and J.H. Bergstrand (2007) Do Free Trade Agreements

Actually Increase Members' International Trade? Journal of

International Economics, 71:1, 72-95.

Bourgheas, S., P.O. Demetriades, and E.L. Morgenroth (1999)

Infrastructure, Transport Costs and Trade. International

Economics, 47:1, 169-189.

Deardoff, A. (1998) “Determinants of Bilateral Trade: Does Gravity

Model Work in a Classical World”? Available at:

http://www.nber.org/chapters/c7818.pdf

Disdier, A.C., L. Tai, L. Fontagné, and T. Mayer (2010) Bilateral Trade

of Cultural Goods. Review of World Economy, 145:4, 575-595.

Duval, Y. and C. Utoktham (2011) “Trade Facilitation in Asia and the

Pacific: Which Policies and Measures Affect Trade Costs the

Most?” Asia-Pacific Research and Training Network on Trade

Working Paper Series, 94. Available at: http://www.unescap-

.org/tid/publication/swp511.pdf Jack


Duval, Y. and C. Utoktham (2011) “Trade Costs in the India-Mekong

Sub-region: Identifying Policy Priorities for Trade Facilitation”.

Working Paper 04/11, Trade and Investment Division, Bangkok:

UNESCAP. Available at:

www.unescap.org/-tid/publication/swp9411.pdf.

Evenett, S.J., and W.K. Hutchinson (2002) The Gravity Equation in

International Economics: Theory and Evidence. Scottish Journal

of Political Economy, 49:5, 489-490.

GOP (various editions) “Pakistan Economic Survey”. Finance Division,

Government of Pakistan.

Hamid, N. and S. Hayat (2012) The Opportunities and Pitfalls of

Pakistan’s Trade with China and Other Neighbours. The Lahore

Journal of Economics, 17, 271-292.

Hoekman, B., and A. Nicita (2011) Trade Policy, Trade Costs, and

Developing Country Trade. World Development, 39:12, 2069-

2079.

Hummels, D.L. (1999) Toward a Geography of Trade Costs. Available

at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=160533

Kayani, F.N., M. Ahmed, T.A. Shah and U.N. Kayani (2013) China-

Pakistan Economic Relations: Lessons for Pakistan. Pakistan

Journal of Commerce and Social Sciences, 7:3, 454-462.

Khan, A. H. and J. Weiss (2006) “Infrastructure for Regional

Cooperation”, Paper Presented at LAEBA, Third Annual

Conference on ‘Regional Integration and Regional Cooperation

in Asia and Latin America: The Role of Regional Infrastructure’

Organised by the Asian Development Bank Institute (ADBI) and

Inter-American Development Bank (IDB), Seoul, 16-17

November.

Leamer, E.E. and J. Levinsohn (1995) “International Trade Theory: The

Evidence”. In G.M. Grossman and K. Rogoff (eds.) Handbook of

International Economics, vol. 3, Amsterdam: Elsevier.

Levin, A., F. Lin and C.J. Chu (2002) Unit Root Tests in Panel Data:

Asymptotic and Finite Sample Properties. Journal of

Econometrics, 108, 1–24.

Lili, X. (2011) “The Factors on Trade of Chinese Cultural Goods: An

Empirical Analysis Based on Panel Data”. School of Economics

http://www.unescap.org/-tid/publication/swp9411.pdf


and Business Administration. Available at: http://www.econ.-

core.hu/file/download/sebawsh/QU_RUXIAO.pdf.

Limao, N. and A.J. Venables (2001) Infrastructure, Geographical

Disadvantage, Transport Costs and Trade. World Bank Economic

Review, 15:3, 451-479.

Lionetti, S. and R. Patuelli (2010) Trading Cultural Goods in the Era of

Digital Piracy. Working paper 40-09.The Remini centre for

economic analysis, Remini, Italy.

Novy, D. (2013) “Gravity Redux: Measuring International Trade Costs

with Panel Data”. CEP Discussion Papers dp1114, Centre for

Economic Performance, LSE.

Olper, A., and V. Raimondi (2008) Explaining National Border Effects

in the QUAD Food Trade. Journal of Agricultural Economics,

59:3, 436-462.

Olper, A., and V. Raimondi (2009) Patterns and Determinants of

International Trade Costs in the Food Industry. Journal of

Agricultural Economics, 60:2, 273-297.

Pedroni, P. (2000) Fully Modified OLS for Heterogeneous Cointegrated

Panels. The Review of Economics and Statistics, 83:4, 727-31.

Saleem, N., S. Jabeen, S. Omer, and M.A. Hanan (2014) Indo-Pakistan

Trade Relations: A Critical Discourse: Analysis of Daily Dawn.

South Asian Studies, 29:1, 309-320.

Saleem, S. and Z. Mahmood (2014) Cultural Goods as a Transformative

Force in the Global Economy: A Case of Pakistan. Available at:

http://www.nust.edu.pk/INSTITUTIONS/Schools/S3H/Researc

h/Documents/Saba%20Salim-3.pdf.

Singh, S. and S.K. Mathur (2014) “Trade Costs of India within Asia:

Measurement and its Determinants”. Working Paper No: 746.

Retrieved from: http://www.freit.org/WorkingPapers/Papers/-

TradePolicyRegional/FREIT746.

WTO (2007) World Trade Report: Trade in a Globalizing World.

Geneva: World Trade Organization.

http://www.econ.-core.hu/file/download/sebawsh/QU_RUXIAO.pdf

http://www.econ.-core.hu/file/download/sebawsh/QU_RUXIAO.pdf

http://ideas.repec.org/s/cep/cepdps.html



http://www.nust.edu.pk/INSTITUTIONS/Schools/S3H/Research/Documents/Saba%20Salim-3.pdf

http://www.nust.edu.pk/INSTITUTIONS/Schools/S3H/Research/Documents/Saba%20Salim-3.pdf

http://www.freit.org/WorkingPapers/Papers/-TradePolicyRegional/FREIT746

http://www.freit.org/WorkingPapers/Papers/-TradePolicyRegional/FREIT746

http://www.freit.org/WorkingPapers/Papers/TradePolicyRegional/FREIT746

BOOK REVIEW

Mundial, B. Poverty and Shared Prosperity 2016: Taking on

Inequality. World Bank Group, Washington DC: Banco Mundial DOI,

10, 978-1. 2016. 193 pages. (Paperback): $39.95.

Poverty and shared prosperity 2016 – Taking on Inequality, is

the first version of a series of annual reports, which is designed and

generated by the World Bank (WB) group. In the book form, the report

is written by several experts from the field of rural development, coming

together as an expert team, thus publishing this comprehensive

manuscript under the name banner of WB as their first edition. It

highlights four key areas; status of poverty through citing actual

statistical figures, shared prosperity and inequality, pressures on the

significance of reduction in inequality and promotion of shared

prosperity throughout the globe. Similarly, utmost importance of

equality along with shared prosperity within a country. The report also

sheds light on the failings and limitations in the capacities of research,

data collection, availability and adequacy of evidence, which hinder the

prosperity and develop causes of poverty and inequality. This report

provides an analysis of comparison among and within the countries, also

evaluates their past state to present state in terms of poverty and

inequality (Hardoon, 2017). Furthermore, how these countries have

developed over a course of time to overcome poverty and promote shared

prosperity is also discussed. The series of reports aims to deliver a set of

rules and examples of policy components and interventions. It helps the

countries to overcome poverty and develop into more equal and

prosperous states.

Chapter 1 discusses the two objectives of WB group, along with

the way to achieve them, i.e., end extreme poverty and promote shared

prosperity. These objectives are very ambitious and are expected to be

met by 2030. The chapter further discusses each objective as a separate

component. The first half discusses poverty; how it is measured by

comparing cost of living and purchasing power parities across countries,

and how it can be monitored against the international poverty line,

BOOK REVIEW 267

standards of basic needs and a proper aggregation of the poverty

calculation. The second half discusses shared prosperity; how it can be

measured by the consumption dispersal and development experienced by

the poorest segment of the country, besides how it can be monitored by

the per capita gross domestic product (GDP). This measure of shared

prosperity helps countries to identify the least wealthy people and also

evaluate the levels of inequality in the income distribution. The chapter

further discourses the range of WB for these goals and how other

organizations and agencies have also adopted similar or somewhat alike

goals to reduce poverty and improve equality, where exists.

Chapter 2 features global poverty. It gives not only precise

figures, but also country-wise comparison of poverty, estimation of the

international, and state personal poverty line and standards against which

different countries evaluate their lowest income consumers. The chapter

explains the multiple characteristics of poverty, conditions of the people

living in poverty and the common lifestyle of the global poor across;

unemployment, over populate households, illiteracy, bad health, im-

proper shelter, WASH, and lack of a stable livelihood, etc. It further

voices how gender affects the poverty graph and what difference women

can make to the household’s income increase. Further, the chapter points

out that the countries with the highest number of poor people and

countries with the highest graph of poverty are often not the same, e.g.,

countries of Middle East, Asia and Africa.

A promotion in the shared prosperity worldwide is reported in

Chapter 3. A total of 83 countries were monitored in 2008-13, out of

which 60 countries (89% world population) have experienced positive

income growth, in spite of the global financial crises of 2008-09 (mb-

KP, 2016). However, the analysis presented in the chapter shows that 23

countries indicated negative income growth. Further, two different

assessments have been made to measure supplementary advancement

and the shared prosperity premium; which compares the relative growth

in income or consumption [Cobham (2016)]. The chapter accounts the

findings that 49 out of these 83 countries showed greater income growth

than the mean (positive shared prosperity premium), while the remaining

34 showed less thriving results (negative shared prosperity premium).

This shows that according to the trend of the Gini coefficient by region,

268 BOOK REVIEW

the highest levels of inequality are observed in Latin America and the

Caribbean. From all the data received and studied, the chapter concludes

that positively shared prosperity premium can lead to the achievement of

the World Bank’s goal of ending extreme poverty by 2030.

Chapter 4 reveals why poverty is reliant on the inequality of

earnings and consumption outgoings. As inequality is the prime focus of

this report, this chapter is also divided in two sections. The first half

stresses upon why and how inequality matters i.e. how it affects poverty

and country’s economic health. The second half discusses global trends

regarding inequality. Different countries define and counter inequality

and equity differently. The chapter provides a cross-country analysis,

which examines income poverty and inequality in terms of sex, health-

care, education, ethnicity, individual or groups. It is to end poverty in a

country through advanced development, reduction in inequality or a

blend of them. The chapter shows that according to the trend of the Gini

coefficient by region, the highest levels of inequality are observed in

Latin America and the Caribbean [PAHO (2017)].

Chapter 5 educates the drivers behind the significant progress in

a shared prosperity, narrowing inequality and reducing poverty achieved

by five countries of the world, i.e., Brazil, Cambodia, Mali, Peru, and

Tanzania. The chapter describes how low to middle income countries

have succeeded in cutting down poverty and inequality – setting an ex-

ample for the world to follow. Each country’s outstanding programmes

are analyzed in detail to measure the ‘whys and wherefores’ that made

these programs successful in the longer run. Concluding this chapter, the

authors come up with a set of options and opportunities that work in the

real world. Also, the drivers for poverty reduction and dismantling

inequality at the country level have been highlighted.

Last two chapters are devoted to policy interventions, which are

based on the understandings and dependable success models from all

around the world. It is to help developing countries to reduce poverty,

enhance shared prosperity and educate developed countries to cut down

inequalities. The chapters advocate the establishment of basic human

rights in the form of healthcare, quality education, food and nutrition,

livelihood opportunities in the form of cash transfers and microcredit

programmes, communication and transport infrastructure development

BOOK REVIEW 269

and advancement of the rural sector as it homes the general poor

community all around the world. The chapters provide rapid strategies

and also long term policy interventions that can contribute in attaining

the universal goal of ending extreme poverty, inequality and sharing

prosperity.

Poverty and Shared Prosperity 2016-Taking on Inequality, is a

successful milestone in the journey to eradicate world poverty. While

reporting the major cause of poverty i.e. inequality, the book opens the

pathways to report other associated key issues of the underdeveloped

societies. It is to discuss in detail that how other issues, e.g., policy

interventions, play an important role in poverty reduction. The report

series will serve as a landmark in development and assist governments,

state collaborations, international and national organizations, private

local bodies and individuals as well. The first report stands as an

impressive work that sheds light over multiple aspects of poverty, around

the world situation, statistics and success models. The report then

delivers need based policy narratives that are definite to be effective in

combating inequality and eradicating world poverty by 2030.

Mahrukh H. Durrani

Post-Graduate Student,

Department of Development Studies,

School of Social Sciences and Humanities (S3H),

National University of Sciences and Technology (NUST),

Islamabad.

Umer Khayyam

Assistant Professor,

Department of Development Studies,

School of Social Sciences and Humanities (S3H),

National University of Sciences and Technology (NUST),

Islamabad.

270 BOOK REVIEW

REFERENCES

Cobham, A., & Sumner, A. (2016). Is It All About the Tails? The Palma

Measure of Income Inequality. Center for Global Development

Working Paper, 343.

Hardoon, D. (2017). Oxfam Report Of 2017 "An Economy for 99%".

Oxfam GB for Oxfam International.

The Dawn. (2016). Key points of the Land Law in Ecuador. (Retrieved

from http://www.thedawn-news.org/2016/01/12/key-points-of-

land-law-in-ecuador/, dated: Oct. 07, 2017))

PAHO. (2017). Values And Principles Of Universal Health. Pan

American Health Organization (Retrieved from paho.org-salud

en las Americas: http://www.paho.org/salud-en-las-americas-

2017/?lang=es&post_t_es=valores-y-principios-de-la-salud-

universal, dated: Oct. 28, 2017)

http://www.paho.org/salud-en-las-americas-2017/?lang=es&post_t_es=valores-y-principios-de-la-salud-universal




SUBSCRIPTION FORM

Kindly enter a subscription of NUST Journal of Social Sciences and

Humanities for the year(s) ……………………………………………....………………………

…………………………………………………..…………………………...................................

.........................................................................................................................................................

In the name of: …………………………………………………………...………………………

.……………………………………………..…………………………………………………..……………………………………………………………………………………………………….

Address: ………………………………………………………………………………………….

…………………………………..……………………………………….…………………………………………………………………………...........................................................................

The Review should be supplied by surface/air mail. A bank draft for the sum of Pak. Rupees/US$..................... is enclosed to cover the above subscription.

Signature: …………………………………………

Date: …………………………………………

Please address your order to: Managing Editor, NUST Journal of School of Social Sciences and

Humanities (S3H), National University of Sciences and Technology (NUST), Sector H-12,

Islamabad, Pakistan

E-mail: [email protected]

Website: http://www.nust.edu.pk/INSTITUTIONS/School/S3H/NJSSH/Pages/Default.aspx

SUBCRIPTION RATE

Inland Annual Per Copy Institutions Rs. 1000.00 Rs. 800.00

Individuals Rs. 800.00 Rs. 600.00

Students Rs. 600.00 Rs. 400.00

Overseas

Institutions US$ 250.00 US$ 100.00

Individuals US$ 200.00 US$ 80.00

Students US$ 150.00 US$ 60.00

Note: Banks Drafts/Pay Orders should be in favour of NUST.


http://www.nust.edu.pk/INSTITUTIONS/School/S3H/NJSSH/Pages/Default.aspx

NJSSH National Referees

Saadia Abid, Quaid-i-Azam University, Islamabad

Rao Nadeem Alam, Quaid-i-Azam University, Islamabad

Fazal Hussain, Pakistan Institute of Development Economics, Islamabad

Asma Hyder, Planning Commision of Pakistan

Muhammad Idrees, Quaid-i-Azam University, Islamabad

Javed Iqbal, Quaid-i-Azam University, Islamabad

Abdul Jalil, Quaid-i-Azam University, Islamabad

M. Jamil Jami, Quaid-i-Azam University, Islamabad

Muhammad Jamil, Quaid-i-Azam University, Islamabad

Attiya Yasmin Javed, Pakistan Institute of Development Economics,

Islamabad

Gulzar Khan, Islamabad Model College for Boys, F-10/4, Islamabad

Muhammad Idrees Khawaja, Pakistan Institute of Development

Economics, Islamabad

Tahir Mahmood, Punjab University, Lahore

Bilal Mirza, University of Central Punjab, Lahore

Shahbaz Nasir, Pakistan Telecommunication Authority, Islamabad

Nasreen Rafiq, F.G. College, F-7/2. Islamabad

Shehnaz Rauf, Air University, Islamabad

Rehahna Siddique, Pakistan Institute of Development Economics,

Islamabad

Shaheena Tariq, Fatima Jinnah Women University, Rawalpindi

Shaista Waqar, Quaid-i-Azam University, Islamabad


GUIDELINES FOR THE SUBMISSION OF MANUSCRIPTS 1. Manuscripts should contain original and unpublished work and must not have been

concurrently submitted to other publishers. It should be materially different from any

published document. This may include significant updating, new analysis or interpretation

of previous work. Maximum manuscript length is 25-30 pages, excluding references and

figures. It should be accompanied by an abstract of not more than 150 words. The first page

of the manuscript should include the name, position, institutional affiliation, postal address

and e-mail addresses of all the authors, indicating who the corresponding author is.

2. Include the following information in an unnumbered footnote appearing on the first page of

the manuscript: names of author(s), titles, affiliated institution(s), addresses, contact

numbers, and e-mail. Include any acknowledgment in this unnumbered footnote. Do not

include this footnote in the consecutive numbering of footnotes. Remaining footnotes (no

end notes) should be numbered consecutively.

3. Manuscripts should follow the publication manual style of American Psychological

Association (APA) sixth edition.

4. Submissions should be in MS Word for the text, in Times New Roman, font size 12 double

spaced and MS Word Equation Editor for equations. Manuscript should carry a margin of

an inch and a half on the left-hand side of the typed page and of at least an inch on each of

the other three sides. The files of tables and charts should be submitted in MS Excel and

stata-generated figures should be saved and sent in Encapsulated PostScript (eps) format.

Each table should a separate set of footnotes.

5. Spell out all abbreviations and acronyms when they are first mentioned in the text. This rule

does not apply for the commonly known and exceptionally long abbreviation. Do not use

abbreviations and acronyms in titles and headings. Abbreviations in tables and figures are

allowed provided these are spelled out in a footnote.

6. Place appendixes at the end of the paper and number them as Appendix 1, Appendix 2, and

so on. Appendixes should carry complete titles.

7. Refer to all graphs, diagrams, and charts as “figures” and number them consecutively in the

text with Arabic numerals. Place all figures on the page where they are first cited.

8. Authors should be careful to ensure accuracy and consistency in the use of mathematical

material. Discussions in the text must be consistent with figures given in tables and

appendixes. In cases where the derivation of formulas has been abbreviated, present the full

derivation on a separate page or as an appendix. Encode formulas using MS Word Equation

Editor.

9. Present all notes as footnotes. Keep footnotes to a minimum, ensuring that they carry

substantive related material. Do not place reference details in the footnotes; rather present

all bibliographic details in a Reference List. Number footnotes consecutively throughout the

text with Arabic numerals. Use a size 9 point for footnotes.

10. Reference lists are strictly required of all submissions. A one-to-one correspondence

between text citations and the bibliography must be observed. The reference list appears at

the end of the main text (after Appendixes). References should carry complete information.

11. Table values should be rounded to one decimal place. Number tables consecutively in the

text using Arabic numerals. Present tables on the page where they are first cited. All tables

should carry the table number and title. Use a size 9 to 11 point within tables, and size 7.5

point for table footnotes.

12. Manuscripts (hard copy) of articles, comments, rejoinders, notes and book reviews in

English only should be sent in triplicate together with a soft copy on CD to the Managing

Editor, NUST Journal of Social Sciences and Humanities, School of Social Sciences and

Humanities, National University of Sciences and Technology (NUST), Sector H-12,

Islamabad, Pakistan. Submissions via e-mail (as attachment in MS Word only) are

acceptable for consideration. Submissions and queries regarding submission may be sent to

[email protected].

13. For further detail please, do visit our website: http://www.nust.edu.pk/INSTITUTIONS/School/S3H/NJSSH/Pages/Default.aspx