Determinants of energy inflation in Pakistan: an empirical analysis.
Haider, Adnan ; Ahmed, Qazi Masood ; Jawed, Zohaib 等
ABSTRACT
This paper aims to identify the key determinants of energy
inflation in Pakistan. The motivation of this study arises due to the
on-going issues related to energy crisis and its price setting
behaviour. The empirical analysis based on OLS, GLS and GMM reveal that
international oil prices, money supply, bilateral exchange rate of
Pak-rupee with US-dollar, tax revenue collection as ratio of
manufacturing sector value added, energy import-gap ratio and adaptive
expectations are critical determinants, which positively contribute to
the energy inflation. The behaviour of monetary and fiscal authorities
seems to be pro-cyclical in response to energy supply side bottlenecks.
This pro-cyclical behaviour along with any international oil price shock
put upward pressure on energy inflation. Be that as it may, the role of
governance and effective long term planning to meet energy demand will
be crucial in order to maintain energy inflation within a single digit
level. Such stance from the government is highly desirable for the
overall welfare of our country, especially for the poor segment in our
society.
JEL Classification: E31, Q41, Q43
Keywords: Energy Inflation, Supply Shortages, Welfare
1. INTRODUCTION
Energy inflation has remained a significant topic in macroeconomic
policy for the past few decades. This is due to several reasons
pertaining to both demand and supply sides. In addition, the history of
energy prices has also been characterised by extreme volatilities,
Hamilton (2008). This makes forecasting and modelling of energy prices
difficult, nevertheless it is important to model and forecast energy
prices in all economies. In this paper we have tried to identify the
determinants of energy inflation in Pakistan.
Energy products are a critical component in any economy, serving as
a core input, particularly in manufacturing industries. Moreover, the
demand for energy and fuel comes from households fuelling cars and
kitchens for which other alternatives are not easily available. This
renders the demand inelastic compared to any other good [Edelstein and
Kilian (2009)], making economies vulnerable to supply and price shocks.
The energy price inflation therefore through cost push inflation and
demand-pull inflation has a major impact on core inflation itself,
thereby playing a significant role in macroeconomic health of a country.
As predicted by Ben Bemanke for the US in- 2006, "in the long run
energy prices can reduce the productive capacity of US economy if high
energy costs make businesses less willing to invest new capital".
The nature of the energy market itself creates a major gap between the
oil consumers and oil producers. Whilst demand is inelastic everywhere,
supply is limited and is difficult to increase, and confined to certain
regions on Earth. This is true particularly for two of the most common
energy types: oil and gasoline. The supply of oil is controlled by a few
countries, and supply shocks therefore lead to an immediate surge in
prices. The oil shock of 1970s created by OPEC caused a major setback
for all oil importing countries of the world, resulting in a global
recession. Energy prices shot up, creating huge demand supply gaps. This
is the case particularly for energy importing countries, which were
helpless in the face of a strong demand and supply gaps. Energy policy
now aims to bridge this gap so that such recessions do not occur again
[Kolev and Riess (2007)].
In an emerging economy like Pakistan, the issue is of prime
concern. With growing industrialisation and high population growth
rates, the demand for energy in Pakistan is set to increase in the
coming years. Pakistan is a net importer of oil, which makes it
vulnerable to oil supply shocks putting subsequent pressure on its
import bills. It has become of critical importance to address the main
causes underlying energy price inflation, and take policy measures to
mitigate such concerns in a timely manner. Energy inflation in Pakistan
is no different than the normal inflation. As we can see in Figure 1,
the trend of energy inflation moves in accordance with the aggregate
inflation. In the past there have been several studies identifying
determinants of the aggregate inflation and more specifically, those
studies were focused on food inflation. Our attempt in this paper is
targeted specifically to study energy inflation.
[FIGURE 1 OMITTED]
In the next section, we will be talking about some stylised facts
about energy. Section 3 will cover the literature review. Section 4 will
discuss the methodology and data. Section 5 will talk about the results
and the last section will conclude the paper along with policy
recommendations.
2. SOME STYLISED FACTS
This section provides some stylised-facts and an overview about
energy outlook both globally and domestically.
2.1. International Energy Outlook
Before analysing domestic energy scenario, some highlights are
obtained from Global Energy Outlook 2012. International energy
consumption pattern is changing globally over the years. According to
International Energy Outlook (IEO) 2012, liquids supply the largest
share of world energy consumption over the projection period, but their
share falls from 34 percent in 2010 to 28 percent in 2040, largely in
response to a reference case scenario in which world oil prices are
expected to remain relatively high. Due to this surge in international
oil prices, use of gas and coal has been gaining importance. Natural
gas, and coal are expected to continue supplying much of the energy used
worldwide, meanwhile, the share of nuclear energy has remained stagnant
(see, Figure 2). Although growth in the energy consumption was 11
percent in 2010, the annual increase of only 1.6 percent would lead to a
15 percent growth in consumption by 2040.
[FIGURE 2 OMITTED]
The IEO 2012 has also projected a steady rise in the demand for oil
in Asia by 2040. This rise in demand is due to higher economic growth
and it is expected that oil consumption of the Asian region will exceed
the North America by 2010; and by 2020 its demand will become nearly
half of the world's total demand for oil. The use of liquids and
other petroleum grows from 87 million barrels oil equivalent per day in
2010 to 97 million barrels per day in 2020 and 115 million barrels per
day in 2040. The liquids share of world energy consumption declines
through 2030, however, as other fuels replace liquids where possible. In
most regions of the world, the role of liquid fuels outside the
transportation sector continues to be eroded. Liquids remain the most
important fuels for transportation, because there are a few alternatives
that can compete widely with liquid fuels. On a global basis, the
transportation sector accounts for 63 percent of the total projected
increase in liquids use from 2010 to 2040, with the industrial sector
accounting for virtually all of the remainder.
[FIGURE 3 OMITTED]
The rising demand for oil and its limited supply has created deep
concern throughout the world as it is believed that nearly all the
largest oil fields have already been discovered and are being exploited.
To meet this demand pressure, total supply in 2040 is projected to be
28.3 million barrels per day higher than the 2005 level of 84.3 million
barrels per day. It is also assumed that OPEC producers will choose to
maintain their market share of world liquids supply, and that OPEC
member countries will invest in incremental production capacity so that
their conventional oil production represents approximately 40 percent of
total global liquids production throughout the projection period.
Increasing volumes of conventional liquids (crude oil and lease
condensates, natural gas plant liquids, and refinery gain) from OPEC
members contribute 13.8 million barrels per day to the total increase in
world liquids production, and conventional liquids supplies from
non-OPEC countries add another 11.5 million barrels per day (see, Figure
3).
2.2. Pakistan's Energy Outlook
Pakistan with a population of more than 180 million has been on the
path of rising GDP growth for the last four years, however a decline in
output growth is observed in FY12 but its growth is still on the higher
side as compared with other developing countries. Real GDP growth
reached 3.4 percent in FY11, after a recovery in FY12, real output
growth reached 7 percent. However, after a moderate decline it dropped
to 5.8 percent in FY08. Since energy sector has a direct link with the
economic development of a country. So energy consumption has also grown
rapidly consistent with the high growth rate of real GDP.
[FIGURE 4 OMITTED]
There has been a consistent energy consumption mix pattern in
Pakistan since FY91. Per capita energy consumption of the country is
estimated at 14 million btn. The energy consumption has grown at an
annual average rate of 4.5 percent from 1990-91 to 2011-12. The major
change in energy mix has taken place in the share of oil and gas
consumption. The share of oil in energy consumption mix has dropped from
48 percent in FY97 to 32 percent in FY11. Figure 4 demonstrates the
energy mix in 2011-12, where oil accounts for 29 percent of the total
energy used.
[FIGURE 5 OMITTED]
Figure 5 (a to d) shows the trend in the use of different sources
of energy in the last ten years. Oil consumption although has declined
in the last few years but still accounts for 29 percent of the total
energy consumed. Consumption of other energy sources shows rising trend
with a moderate decline in last fiscal year. When we talk about energy
inflation, the first variable, which comes to our mind is international
oil price. What affect does oil price have on energy inflation? An
increase in oil price is expected to have a reduction in standard of
living by 20 percent for the oil importing countries and vice versa for
the oil exporting countries [Thoresen (1982)]. If the direct impact does
not reflect itself in wage reduction, it will be evident through high
inflation. Although in some countries, the inflationary effect of oil
price is limited despite the fact that fluctuations in crude oil price
is a key element in inflation variation [Alvarez, et al. (2011)]. In
Pakistan, we experience an indirect effect of Oil price. Figure 6 shows
the overall trend of energy inflation in Pakistan and oil prices for the
period of 1991-2012.
The fact that inflationary effect of oil price is weak can be
witnessed from Figure 6. During FY08, a spike in oil prices was
witnessed due to the global financial crises but on the other hand
energy inflation in Pakistan has been consistent and showed a smooth
behaviour. Nevertheless, we have been experiencing a rise in fuel and
electricity charges over the years. This graph also shows the prices of
crude oil, gas, firewood, electricity and kerosene-oil for the period
1991-2012. We can see that prices of different commodities have been
growing at different exponential rates with different volatilities over
time.
[FIGURE 6 OMITTED]
The crude oil prices have increased significantly over time,
showing high volatility. There have been cyclical rises and falls, with
prices peaking during the International Oil Shock during mid-2008. After
facing a huge downfall right after the price shock, the prices soon rose
to high levels in 2012. The rise in gas prices is relatively less sharp
than crude oil, showing little volatility in the earlier periods, and
increasing volatility in the later periods. It is also notable that the
crude oil prices have been rising and falling about the steady trend of
gas prices, such that the average prices of both commodities appear to
be fairly similar. Only in the 2008 Oil Shock, the crude oil prices
drastically differed from the gas prices. The firewood prices too have
been rising steadily with no volatility. The prices of kerosene oil rose
very slowly in the earlier periods but began to show higher growth rates
after 2008. The electricity prices were fairly stable and experienced a
slow growth in comparison to the prices of other commodities. It is
evident however, that over time there has been a rise in all types of
commodity prices, contributing to overall energy price inflation.
However, these sudden increments in prices lead us to investigate the
determinants of energy inflation in the case of Pakistan.
3. SELECTED LITERATURE REVIEW
A low and stable level of inflation is one the major goals of any
economy. The question then arises that how to achieve a low level of
inflation or how to maintain inflation at the current level. To
understand this, one needs to look at the causes of inflation. Inflation
can be caused in two broad situations. One is where "too much money
is chasing too few goods" i.e., demand pull inflation. The other is
when increase in prices of raw materials drive up costs of production,
which feed into the prices of finished goods. This is referred to as
cost push inflation. Inflation has a twofold effect on the economy. It
can be bad as well as good. There is a threshold level beyond, which
inflation can be harmful to the economy [Bruno and Easterly (1998); Khan
and Senhadji (2001); David, et al. (2005)]. In the case of Pakistan that
threshold happens to be 9 percent [Mubarik (2005)]. However, Hussain
(2005) suggests a 3 percent--6 percent inflation rate to have positive
effects on Pakistan's economy. It provides incentives to
production, investment and growth in wages. Friedman (1970) presented
the theoretical foundations on the quantity theory of money, which is
the part of classical economic theory. He argued that "inflation is
always and everywhere a monetary phenomenon". Friedman and Schwartz
(1970) tested it empirically. The classical are of the views that
increase in the money supply results in proportionate increase in
prices, assuming economics agents are rational and output and real money
balances are constant.
In the context of Pakistan, the history for analysing the
determinants of inflation started 30 years ago when [Khan (1982)]
concluded that demand for money improves the variation in the rate of
inflation. Till 1989, inflation has been seen as a monetary phenomenon.
Saleem (2008) also suggested the same. Her argument was inflation,
interest rate and money supply move in the same direction.
A look at recent inflation trends in Pakistan helps give a snapshot
of inflation as well as factors affecting inflation as a whole. [Khan,
et al. (2007)] endorse a dynamic approach to determining causes of
recent inflation in Pakistan in 2005-06. High growth rates were also
accompanied with sharp rises in inflation. Keeping in context the
volatile economies of developing countries they apply a structuralist
approach, which includes both demand and supply side factors. They find
that adaptive expectations have been one of the key determinants of
inflation in Pakistan over the decades as well as in the period of
2005-06. This is through the channel of food prices as more than half of
the budget of the poor comprises of food expenditure. Overall for the
year 2005-06, the adaptive expectations contributed 3.66 percentage
points to the inflation rate of 8 percent, explaining 45.73 percent of
headline inflation rate. Non-government sector borrowing was the second
largest contributor (which explained 35 percent of headline inflation or
contributed 2.8 percentage points to the inflation rate of 8 percent).
Knowing inflation trends in general is not enough [Khan and
Schimmelpfennig (2006)] rightly point out that determining what causes
inflation will determine which policy makers are to tackle it. If
inflation is a monetary phenomenon then it is appropriate for the
Central Bank to control it. However if inflation is affected by supply
side factors, and here they look at support prices of wheat, then it
becomes more appropriate for the Ministry of Agriculture to devise a
course of action to deal with inflation. Focusing on headline inflation
and using monthly data, they find that wheat support prices affect
inflation only in the short run, whereas monetary variables of broad
money and private sector credit affect inflation in the long run and by
a lag of 12 months.
Refining the argument further, determining which factors affect
which type of inflation will also determine which policy makers are most
appropriate to deal with the situation. Not much work has been done to
identify the determinants of energy inflation in Pakistan. From the
international perspective, a recent study in Finland [Irz, et al.
(2011)] about determinants of food inflation with its linkages with
energy inflation was conducted. A long run relation was evident between
food inflation and energy prices as well as some other agriculture
products. Similarly, energy inflation is itself a determinant of various
other factors e.g. house prices etc. The correlation was found in a
study conducted for Euro Area. In U.S, it was also evident that energy
prices are a key determinant of inflation [Dhakal, et al. (1994)].
4. DATA AND EMPIRICAL METHODOLOGY
We have used data from 1973-2012 on an annual frequency basis. It
is taken from multiple sources including State Bank of Pakistan, World
Development Indicators, Pakistan Bureau of Statistics. Table 1 provides
list of sources' along with descriptive statistics of selected
variables.
Given the dynamics of our country and past literature, we have
estimated the following equation:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
In this functional form, [P.sup.Energy.sub.t] is energy price
inflation. Other variables, Oil Prices, Money Supply, Nominal Exchange
Rate, Energy Import-Gap Ratio (EIMPR), Govt. Tax Revenue as a ratio of
Value Added to Manufacturing Sector and Adaptive Expectation are used to
determine the energy inflation. [[epsilon].sub.t] represents residual
term. All the variables except Energy Imports-Gap ratio and Tax Revenue
as ratio of Manufacturing Sector are taken in logarithmic form. To
estimate the model, we have used Ordinary Least Square (OLS),
Generalised Least Square (GLS) and Generalised Method of Movement (GMM)
methods. EIMPR is the ratio of energy imports and energy gap that is
prevailing in our country. The rationale behind this key variable is
that it will reflect the significance of energy shortages on prices.
Given the ground realities of our country we expect it to have a
positive sign. Tax revenue as the ratio of the manufacturing sector
value added has also been included as one of the independent variables.
The main rationale behind this variable is that most of the industries
get affected due to the rising electricity prices. Due to energy
shortages, government tends to raise taxes and thus it ought to affect
the energy prices. Any surge in International Oil prices or any
depreciation in the domestic Exchange Rate affect the Import Bill
significantly, which ultimately put pressure on energy inflation. The
temporal relationships of these variables with energy inflation are
shown in Scatter plots available in Appendix. Constantly increasing
prices can create expectations for future inflation. The role of
expectations is critical in analysing future prices. With the increase
in prices, individual expects higher salaries, speculation in asset
prices increases, credit diverts to real estate and stock markets etc.
and rent seekers become active in expecting a higher price in future. A
similar pattern of expectations goes for energy inflation therefore to
incorporate all of these elements we have used a variable of lag of
energy inflation. We expect it to show a positive sign as well.
5. RESULTS AND DISCUSSIONS
This section discusses the results based on our estimation. The
first task was to diagnose the stationary properties of individual
variable. For this purpose we have tested the variables for unit root
using Augmented Dicky-Fuller and Philips-Perron tests. As shown in Table
2, all the variables appear to have an integration of order 1.
After diagnosing the stationarity properties we have estimated the
model using three estimation approaches, OLS, GLS and GMM. Our prime
task is to analyse the long run relationship. For this purpose we have
carried out the co-integration tests. Both Engle-Granger and
Johansen-Juselius show the existence of co-integration vector.
Considering the co-integration results, we estimated our model at level.
The estimation results in Table 3 are promising and show theoretically
correct signs of the coefficients. Broad money, Oil prices, Exchange
rate, Tax Revenue as ratio of Value Added of Manufacturing Sector and
dummy for 2008, are statistically significant. Since the variables are
in log form, the estimated coefficients can be termed as elasticities.
For example, a 10 percent change in Tax Revenue as ratio of Value Added
to Manufacturing Sector can in turn affect the energy inflation by 2.6
percent. Oil prices have an indirect effect on the energy inflation in
our country. It is significant with a lag of one year. The coefficient
of 0.082 does not show a large impact but the t-stats of 4.52 make it
quite a significant variable. We have discussed immense literature that
talks about inflation being a monetary phenomenon and therefore Broad
Money has always been a significant variable in determining the
inflation. In our results, it is highly significant with t-stats of 9,8.
The coefficient of 0.34 can be interpreted as when there is a 10 percent
increase in money supply, it will affect energy prices by 3.4 percent.
Energy Import-Gap ratio shows a positive sign but it is insignificant.
Primarily because it is a ratio and the quantum of energy imports is
low. Exchange Rate is another highly significant variable with t-stats
of 5.74 and with coefficient of 0.38. In Pakistan, exchange rate plays a
vital role in affecting inflation. It has the pass through effect
through currency depreciation, which is quite evident from our results.
After adaptive expectations, exchange rate is the most critical variable
that affects energy prices given its elasticity.
The effect of inertia cannot be ignored when we analyse inflation,
especially in Pakistan. Based upon the estimation results, we have
calculated the contribution of the explanatory variables in explaining
energy inflation. In Pakistan, Adaptive Expectations have always been a
major contributory factor in explaining inflation, food as well as
energy. People in Pakistan, expect prices to grow rather than decline.
As expected, the significance of Adaptive expectations is quite evident.
Over the last 30 years, it has contributed almost 42 percent on an
average to the total energy inflation. The second contributing factor is
the broad money with an average contribution of 37 percent over the same
period. During the early 70s, energy inflation was 15.3 percent, highest
ever in the past 30 years. It was mainly because of the 70s oil price
shock, which created an expectation of high inflation. This can be seen
since adaptive expectations are contributing almost 42 percent with a
value of 6.5 percent in the overall inflation of 15.3 percent.
[FIGURE 7 OMITTED]
Early 80s and 90s are the era of low inflation. Private sector
borrowing, broad money and adaptive expectations were the main factors
contributing to this energy price growth. Contribution during the 80s
and 90s remained consistent to 40 percent as far as inertia is
concerned. As for exchange rate, its contribution surged from 18 percent
in early 80s to 29 percent in the late 90s. Such an increase was
probably due to the frequent change in governments and inconsistency of
policies, nuclear explosion and other political uncertainities. During
this time period, energy inflation was around 8.2 percent. In early
2000, energy inflation declined to 4.5 percent and broad money was the
major contributor to it, explaining almost 60 percent of energy
inflation. During 2006, inflation shot up again to 10.4 percent,
adaptive expectations alone explained almost 40 percent of it. It was
considered to be the highest compared to the average inflation of 4.5
percent during the millennium decade. The major hike in inflation during
this time was probably due to the surge in house rents. In 2008-09
energy inflation took a sudden bump to 18 percent due to the oil price
shock. However the impact was not significant enough to last for a
longer period. This may be probably because our economy is not directly
affected by international oil prices. By 2011-12, energy inflation came
back to 11 percent with adaptive expectation contributing 6.4 percent
(60 percent of energy inflation). It was because of the oil price shock
that people were expecting that the effects of the shock will last for
coming years or so. But in 2012 this expectation dropped by 4 percent
explaining 36 percent of energy inflation. Other factors such as Nominal
Exchange rate and broad money were other major contributors with 2.2
percent (explaining 20 percent of energy inflation) and 2.7 percent (25
percent of energy inflation) respectively.
The second most important factor was broad money. Broad money over
the period of last three decade has contributed around 3 percent (38
percent of energy inflation) on an average to the energy inflation
according to our finding. In 80s and 90s broad money contributed around
3.1 percent (37 percent if energy inflation) on an average. It was the
second major contributor after adaptive expectations. During the first
half decade of 2001-05 its contribution shot up to 2.8 percent (60
percent of energy inflation) when energy inflation was 4.6 percent and
growing. From 2006-2012, energy inflation has been growing and reached
11 percent and contribution of broad money has declined to 2.7 percent
(25 percent of energy inflation). It was mainly because other factors
such has house prices, oil prices came into play that resulted in high
energy inflation. Nominal Exchange rate contributed 1.6 percent (16.5
percent of energy inflation) over the last 3 decades. Its highest
contribution was witnessed in late 90s when it rose to 2.5 percent (30
percent of energy inflation) after adaptive expectations. This may have
occured due to the uncertain changes in governments and frequent changes
in policies, nuclear explosion and other political situation that was
prevailing during that era.
The above mentioned factors were not the sole contributors to
energy inflation. By 2008-09 the issue of circular debt came into play,
which raised the electricity prices to a great extent. The period of
2008-12 was a period of power sector inefficiencies, which have cost the
country significantly directly in the form of budget costs in the last
five years.
6. CONCLUSION
This paper talks about the determinants of energy inflation in
Pakistan. It evaluates the role of various factors in explaining energy
inflation, which include broad money, exchange rate, international oil
prices and adaptive expectations. These are the most prominent factors
in explaining energy inflation. Others includes tax revenue as a ratio
of the manufacturing sector value added and energy import-gap ratio. Our
results conclude that the behaviour of monetary (State Bank of Pakistan)
and fiscal (Ministry of Finance) authorities seems to be pro-cyclical in
response to energy supply side bottlenecks. This pro-cyclical behaviour
along with any international oil price shock and exchange rate
depreciation put upward pressure on energy inflation. The heavy reliance
of government on indirect taxation of energy items strongly hits the
poor segment of the society. Supreme-court of Pakistan has also observed
this behaviour of the government, which reduces the overall welfare. As
expected approximately 60 percent of the energy inflation is explained
by adaptive expectations. Given such issues including that of circular
debt, it is only logical to expect the prices to go up.
Also, there are certain shortcomings in the infrastructure that
need to be taken care of. There is a complete disconnect among all the
stake holders involved at the policy level. The Ministry of Water and
Power being the critical stakeholder does not have a roadmap for itself.
It is more reactive than proactive to the power sector reforms, which is
perceptibly due to lack of political will to improve the system. There
is lack of professional attitude at the regulatory level. The regulator
fails to address the problems of the power sector and is working in
isolation. Furthermore there is no governance at the entities level.
Despite the fact that they are being micro-managed by policy makers and
regulators, they themselves have no attitude of moving forward and are
satisfied in maintaining a status quo. Until and unless there is a
serious accountability for the above mentioned issues, inflation
expectations will always play a significant role. There must be proper
structural reforms that require effort from all. There should be a
roadmap outlined and roles defined for all the entities so that there is
no ambiguity in achieving the ultimate objective.
Adnan Haider <ahaider@iba.edu.pk> is Assistant Professor of
Economics and Finance at Department of Economics and Finance at
Institute of Business Administration, Karachi. Qazi Masood Ahmed
<qmasood@iba.edu.pk> is Professor/Director Research at Center for
Business and Economics Research, Institute of Business Administration,
Karachi. Zohaib Jawed <zjawed@iba.edu.pk> is Research Assistant at
Center for Business and Economics Research, Institute of Business
Administration, Karachi, Pakistan.
Authors' Note: We would like thank to Wasim Shahid Malik (for
providing us useful comments as discussant), M. Ather Elahi and Mehwish
Ghulam Ali for their fruitful discussions on this topic. Any errors or
omissions in this paper are responsibility of the authors. Views
expressed here are those of the author and not necessarily of the
Institute of Business Administration, Karachi.
APPENDIX
Relationships of Energy in Inflation with Key Determinants
[GRAPHIC OMITTED]
REFERENCES
Alvarez, L. J., S. Hurtado, I. Sanchez, and C. Thomas (2011) The
Impact of Oil Price Changes on Spanish and Euro Area Consumer Price
Inflation. Economic Modelling 28, 422-431.
Bruno, M. and W. Easterly (1998) Inflation Crises and Long-run
Growth. Journal of Monetary Economics 41, 3-26.
David, D., G.-P. Pedro, and H. E. Paula (2005) Threshold Effects in
the Relationship between Inflation and Growth: A New Panel-data
Approach. (MPRA Working Paper No. 38225).
Dhakal, D., M. Kandil, S. C. Sharma, and P. B. Trescott (1994)
Determinants of the Inflation Rate in the United States: A VAR
Investigation. The Quarterly Review of Economics and Finance 34, 95-112.
Edelstein, P. and L. Kilian (2009) How Sensitive are Consumer
Expenditures to Retail Energy Prices? Journal of Monetary Economics 56,
766-779
Friedman, M. (1970) A Theoretical Framework for Monetary Analysis.
The Journal of Political Economy 78, 193-238.
Friedman, M. and A. Schwartz (1970) Monetary Statistics of the
United States: Estimates, Sources and Methods. NBER Books
Hamilton, J. D. (2008) Understanding Crude Oil Prices. National
Bureau of Economic Research.
Hussain, M. (2005) Inflation and Growth: Estimation of Threshold
Point for Pakistan. Pakistan Business Review 1-15.
Irz, X., J. Niemi, and Liu (2011) Determinants of Food Price
Inflation in Finland--The Role of Rnergy. MTT. Finland. (Working Paper).
Khan, A. A., S. K. H. Bukhari, and Q. M. Ahmed (2007) Determinants
of Recent Inflation in Pakistan. (MPRA Working Paper No. 16254).
Khan, A. H. (1982) The Demand for Money and the Variability of the
Rate of Inflation: An Empirical Note. Economics Letters 10, 257-261.
Khan, M. and A. Schimmelpfennig (2006) Inflation in Pakistan: Money
or Wheat? The Pakistan Development Review 45, 185-202.
Khan, M. S. and A. S. Snhadji (2001) Threshold Effects in the
Relationship between Inflation and Growth. IMF Staff Papers 1-21.
Kolev, A. and A. Riess (2007) Energy: Revival of a Burning Matter.
EIB Papers 12, 10-28.
Mubarik, Y. A. (2005) Inflation and Growth: An Estimate of the
Threshold Level of Inflation in Pakistan. SBP Research Bulletin 1,
35-44.
Saleem, N. (2008) Measuring Volatility of Inflation in Pakistan.
Lahore Journal of Economics 13, 99-128.
Thoresen, P. E. (1982) Oil Price and Inflation. Energy Economics A,
121-126.
Table 1
List of Variables with Descriptive Statistics
Energy Growth Exchange
Variables Inflation in MS Rate
Data Sources PBS SBP SBP
Mean 9.25 15.68 36.01
Median 8.06 14.85 26.21
Maximum 27.40 26.19 94.42
Minimum 3.28 6.15 9.90
Std. Dev. 5.20 5.15 25.85
Skewness 1.97 0.37 0.67
Kurtosis 7.46 2.58 2.19
Jarque-Bera 57.45 1.18 4.08
Probability -- 0.55 0.13
Sum 360.78 611.44 1440.48
Sum Sq. Dev. 1028.19 1006.65 26054.73
Observations 39.00 39.00 40.00
Energy Tax Revenue
Import-Gap of Manufacture
Variables Oil Prices Ratio Sector
Data Sources SBP WDI PES
Mean 32.35 9.72 0.82
Median 23.66 9.65 0.87
Maximum 108.88 12.17 1.04
Minimum 3.27 8.06 0.54
Std. Dev. 25.58 0.98 0.15
Skewness 1.58 0.52 -0.60
Kurtosis 4.58 2.81 2.09
Jarque-Bera 20.77 1.87 3.82
Probability 0.00 0.39 0.15
Sum 1293.92 388.84 32.75
Sum Sq. Dev. 25519.71 37.12 0.85
Observations 40.00 40.00 40.00
Note: * PBS = Pakistan Bureaus of Statistics.
* SBP = State Bank of Pakistan.
* WDI = World Development Indicators.
* PES = Pakistan Economic Survey.
Table 2
Stationarity Diagnostics: Tests of Unit Roots
Augmented Dicky-Fuller Test
Level Difference Order of
Variables Integration
Energy Inflation 0.974 0.002 I(1)
Money Supply 0.560 0.002 I(1)
Exchange Rate 0.991 0.000 I(1)
Oil Prices 0.984 0.000 I(1)
Energy Import-Gap ratio 0.761 0.001 I(1)
Tax Revenue as Ratio
of Manf Sector 0.846 0.032 I(1)
Phillips-Perron Test
Level Difference Order of
Variables Integration
Energy Inflation 0.463 0.002 I(1)
Money Supply 0.632 0.002 I(1)
Exchange Rate 0.986 0.000 I(1)
Oil Prices 0.997 0.000 I(1)
Energy Import-Gap ratio 0.124 0.000 I(1)
Tax Revenue as Ratio
of Manf Sector 0.819 0.000 I(I)
Note: Authors' Calculations.
Table 3
Estimation Results (Dependent Variable CPI-Energy Inflation)
Sample (Adjusted): OLS GLS
1973-2012
Variables Coefficients t-Stats Coefficients t-Stats
Constant -2.333 -7.75 -0.961 -3.236
Lagged Oil Price 0.082 4.568 0.051 3.005
Broad Money 0.341 9.817 0.128 3.348
Govt. Taxes as a
Ratio of
Manufacturing
Sector 0.260 2.580 0.149 2.117
Exchange Rate 0.387 5.741 0.274 4.755
Energy Import-Gap
Ratio 0.005 0.687 0.011 1.790
Dummy Variable for
2008 -0.057 -2.284 -0.015 -2.316
Adaptive
Expectations 0.668 4.880 0.497 5.473
R-squared 0.999 0.999
Adjusted R-squared 0.999 0.999
Durbin-Watson Stat 1.425 1.592
J-Statistic * -- --
Cointegration
(Engle-Granger) Yes Yes
Cointegration
(Johansen and
Juselius) Yes Yes
Sample (Adjusted): GMM
1973-2012
Variables Coefficients t-Stats
Constant -0.961 -2.885
Lagged Oil Price 0.051 2.680
Broad Money 0.128 2.985
Govt. Taxes as a
Ratio of
Manufacturing
Sector 0.149 1.887
Exchange Rate 0.274 4.239
Energy Import-Gap
Ratio 0.011 1.595
Dummy Variable for
2008 -0.015 -2.117
Adaptive
Expectations 0.497 4.879
R-squared 0.999
Adjusted R-squared 0.998
Durbin-Watson Stat 1.618
J-Statistic * 0.037
Cointegration
(Engle-Granger) Yes
Cointegration
(Johansen and
Juselius) Yes
* Instruments list:= lag terms and lag difference terms.
