The Use The Annual Data

Published Date: 02 Nov 2017

ABSTRACT

Corruption is a widespread problem in Pakistan. Few studies have focused on the determinants of corruption in Pakistan. But I did not find any study carrying empirical work on economic determinants of corruption in Pakistan. This study will empirically investigate the economic determinants of corruption and will also explore the causal relationship between corruption and its economic determinants in Pakistan by using time series data from 1984-2012. The study will also examine the past behavior of corruption in our country. Economic determinants of corruption include Income level, Inflation rate, foreign direct investment, Population growth and Economic competition. To investigate the impact of economic determinants of corruption in Pakistan ARDL (Autoregressive Distributed Lag) technique will be applied. Granger causality test will be applied to check the causal relationship between corruption and its economic determinants in Pakistan.

INTRODUCTION

In the words of Quaid-e-Azam:

"One of the biggest curses from which Sub-continent is suffering, I do not say that other countries are free from it, but, I think our condition is much worse, is bribery and corruption. That really is a poison. We must put it down with an iron hand" [1] 

Corruption is a worldwide problem. It occurs both in developed and under developed countries. Corruption is a menace for any society, as it has a number of negative impacts on the society. It discourages investment, increases crime rate and unemployment rate. According to Brempong the country in which corruption is higher, there is a high Gini coefficient and less government expenditures on health and education sector [O'Rourke 1993]. It means corruption also leads to the higher income inequality and due to which poverty level also rises in a country. Similarly, low expenditures on health and education sector leads to lower human capital (proxied by years of schooling or good health). Thus, corruption has negative impact on human capital in a country. Thus it can be said that Economic growth of a country or society is badly affected by corrupt practices. According to Jain for the existence of corruption three conditions should be fulfilled.1) discretionary power 2) economic rent 3) weak judicial system.

Corruption is worse than prostitution. The latter endanger the morals of an individual; the former invariably endangers the morals of the entire country [Karl Kraus] [2] . According to Transparency International "Corruption is the abuse of entrusted power for private gain. It hurts everyone who depends on the integrity of a people in a position of authority. [3] "

In the "Transparency International Corruption Perception Index" ranging from 0 (most corrupt) to 10 (lest corrupt), Pakistan has consistently maintained a poor ranking. In 2009, 2010 and 2011 Pakistan’s score out of 10 was 2.5, 2.3 and 2.5 respectively. In 2012 out of 176 countries Pakistan was ranked as the 139th most corrupt country of the world and on a scale of 0 (highly corrupt) to 100 (very clean) Pakistan’s score was 27. While the Denmark, Finland and New Zealand were at first place with 90 scores. [4] Thus their score shows that these are the least corrupt countries.

While briefing the media Sohail Muzaffar (the advocate of transparency international) declared that in the rule of law index 2012 out of 97 countries Pakistan ranked as 7th most corrupt country. In 2011 Pakistan was 42nd most corrupt country of the world while in 2012 its ranked increased to the 33rd most corrupt country of the world. This shows that corruption is increasing day-by-day in Pakistan. [5] 

As we know that corruption is a very serious problem that our country is facing but unfortunately this issue has given much less attention in terms of empirical research work than it actually deserves. In Pakistan only few studies have been conducted to investigate the impact of economic determinants of corruption in Pakistan or the causal relationship between economic determinants of corruption and the level of corruption in the country. Zakaria (2009) empirically examined the relation between trade openness and corruption in Pakistan. He found the negative impact of trade openness on corruption. The most recent studies "corruption, causes and effect in Pakistan’s case (A Review Research)" conducted by Khan et al (2012) &"corruption causes and cure" conducted by Haq (2012) were just theoretical. They did not empirically draw their results. So, there is a gap in the literature and to fill this gap there is a need of an empirical work on economic determinants of corruption regarding Pakistan.

This study is an effort made to investigate the impact of economic determinants of corruption in Pakistan by using time series analysis. The economic determinants included in the study are: income level, inflation rate, foreign direct investment, population growth and economic competition. The study uses the time series analysis for all the variables for the period (1984-2012) in Pakistan. The basic purpose of this study is to investigate the proposition as to how economic determinants of corruption influence the corruption level in the country? And also explores the causal relationship between corruption and its economic determinants. The study also aims to examine the past behavior of corruption in Pakistan. According to my knowledge this study is going to be the first study to address this issue in a rigorous way in Pakistan.

OBJECTIVE OF STUDY

To determine the impact of economic determinants of corruption in Pakistan i.e. to what extent economic determinants affect the corruption level.

To investigate the causal relation between corruption and its economic determinants in Pakistan.

To study the past history of corruption in Pakistan.

HYPOTHESIS

Following hypothesis will be tested in the course of study:

H1: Higher level of income is related to the perceived level of corruption.

H2: The level of corruption is correlated with higher inflation.

H3: High foreign direct investment is correlated with level of corruption.

H4: Higher population growth affects the level of corruption.

H5: Economic competition is related to the level of corruption.

LITERATURE REVIEW

Ali and Isse (2003) examined the determinants of corruption by a cross country comparison. The study used education, judicial efficiency, the size of government, political and economic freedom, foreign aid, ethnicity, and the type of the political regime as the determinants of corruption. The study employed OLS regression analysis and granger causality test. The results showed that corruption was found to be negatively and significantly correlated with the level of education, judicial efficiency, and economic freedom. It was positively and significantly correlated with foreign aid and the size of government. While the result of granger causality test implies the one way causality between corruption and GDP growth rate i.e. corruption granger causes the GDP growth rate.

Ata & Arvas (2011) examined the economic determinants of corruption. The study used the cross-section data (2004-2007) including corruption, economic development, growth, inflation, economic freedom and income distribution on 25 member countries of EU. It applied Tobit model of censored regression model and results show that economic development, economic freedom and growth rate is positively related to corruption while inflation and Gini coefficient is inversely related to corruption.

Frechette (2006) examined the panel data analysis of the time-varying determinants of corruption. He took schooling, income level and political freedom as the determinants of corruption. He used the data of 135 countries for 16 years. He applied OLS regression analysis. The results suggests that increase in schooling and income level increases the corruption level while greater political freedom reduces the corruption level.

Haq (2012) theoretically examined the causes and cures of corruption in Pakistan. According to the study the main causes of corruption were corrupt leaders, political instability and interference, weak judiciary system, inefficient governments and lack of transparency. Study also theoretically explained the impact and cures of corruption.

Khan et al (2012) has investigated the causes and effects of corruption in Pakistan through a review research. The study discussed the histort, reasons, forms, issues, consequences, facts & amount and prevention of corruption in Pakistan. According to the study the reasons of corruption in Pakistan were land awards, bloated public sector, private sector cooperatives, drug money, foreign aid and investment, public sector expenditures, underground economy and tax regime. Similarly the issues of corruption were excessive aid flows, loan default, weak regulatory system and rent seeking in land acquisition. But this work was just theoretical and it did not empirically investigate the impact of corruption regarding Pakistan.

O'Rourke (1993) conducted a survey on the determinants of corruption. This study measured the corruption on the basis of: money metric term, micro data and perception of the likelihood, frequency, or level of corruption of respondents in surveys. It divided the determinants of corruption in four categories i.e. economic and demographic determinants, political institutions, judicial and bureaucracy, and geography and culture. It explained the relationship between corruption and each determinant by their expected signs. Income proxied by GDP per capita (+), income distribution proxied by Gini coefficient (+), size of government (+), share of imports in GDP (-) , restrictions on foreign trade (+), foreign investment (+) and capital market (+), economic openness (-), economic freedom (-) measured by indexes of heritage foundation and the wall street journal, or the fraser institute, inflation (+), foreign aid (+), human capital proxied by schooling (-), population growth (+) and structural reforms (-) were taken as economic and demographic determinants. Political institutions include political freedom (-), press freedom (-), unstable policies (-), presidential system (+) and number of political parties (+). Judicial and bureaucratic environment involves quality of judicial system (-), bureaucratic environment (-), increase in public sector wages (-) and decentralization (-). While in geography and culture determinants ethno linguistic homogeneity (-), countries with many protestants (-), colonial heritage (+), women participation (-), countries away from equator (-) and natural endowment (+).

Rehman and Naveed (2007) investigated the determinants of corruption and its relation with growth. For this analysis the cross sectional data of 104 countries for the year 1995-2005 was used. Corruption was used as a dependent variable while real GDP per capita, secondary school enrolment, public spending on education, FDI, and unemployment rate as explanatory variables. The objective was to find out the main determinant of corruption and also to investigate the impact of corruption on growth. Barro cross country regression was applied and the results suggest that explanatory variables are affecting the corruption level and by combining these variables level of corruption can be changed.

The study of Seldadyo and Haan (2006) also examined 70 determinants of corruption. According to this study some important economic determinants and their expected relationship are as follow: income proxied by GDP per capita (+), income distribution proxied by Gini coefficient (+) and size of government (+). Some political determinants are democracy (-), political liberty (-) and decentralization (+). Some bureaucratic and regulatory determinants include quality of judicial system (-), rules of law (-) and quality of bureaucracy (-). Similarly some geographical, religious and cultural determinants are: number of protestants (-), ethno linguistic homogeneity (-) and colonial heritage (+).

Shabbir and Anwar (2007) investigated the impact of various economic and non economic determinants of corruption. The study used cross sectional data for 41 countries. The economic determinants include economic freedom, globalization, education level, level of development and income distribution. While non- economic determinants include democracy, press freedom, share of population having affiliation with particular region. Study concludes that except the distribution of income all economic determinants are negatively related to the level of corruption. While the non- economic determinants are insignificantly explaining the variation in corruption level. Thus economic determinants are more important than the non-economic determinants in determining the level of corruption in developing countries. So government should focus on economic determinants to remove the corruption level.

Tanzi (1998) explained the causes, consequences, scopes and cures of corruption around the world. This study discussed many causes or determinants, effects and corrective measures to reduce corruption. According to the study corruption is closely linked to the way government conduct its affairs. So if government wants to reduce corruption in the society it has to modify the way it operates. And corruption cannot be reduced by single actions. As it is a costly effort, the societies should be together to reduce or eliminate the corruption.

Treisman (1998) examined the causes of corruption by a cross national study. The multiple regression analysis was applied to estimate the causes of corruption. The results of study show that there is an inverse relation between economic development and corruption. To test the robustness he applied Edward Leamer’s extreme bounds analysis. It also concludes that more developed and British colonial countries are less corrupt, while the countries with federal structure are more corrupt.

Ullah & Ahmed (2007) investigated the relation between income inequality and corruption by using panel data from 1984-2002 for 71 countries (developing and developed). This work examined the impact of corruption on income inequality. For this purpose some control variables i.e. education, trade, capital per worker, government expenditure, and population growth was also incorporated . The result of study shows that there is a negative relationship between income equality and corruption i.e. if there is a 1 standard deviation worsening in corruption index it will leads to the 1.3 percentage increase in the Gini coefficient.

Zakaria (2009) empirically examined the relationship between trade openness and corruption level in Pakistan. For this purpose the time series data from 1984-2007 had been used. Corruption was taken as dependent variable while trade openness as an independent variable. He also incorporated some other control/explanatory variables i.e. human capital, government consumption, population, inflation rate, capital stock per worker and defense expenditures. The study hypothesized the negative relationship between corruption and trade openness. By applying OLS regression analysis the results show that trade openness has significant negative impact on the corruption. With an increase in educated population and inflation rate corruption level reduces. Similarly, an economy with larger government, increase in population, military spending and capital population breeds the corruption level.

In the above studies the researchers mostly used cross-sectional data for cross-country analysis in different time periods or just theoretically explained the determinants of corruption in Pakistan. In Pakistan more attention has been paid on theoretical work on the causes of corruption. Even the most recent researches on determinants of corruption in Pakistan "Corruption: Causes and Effects in Pakistan’s Case (A Review Research)" by Dr. Khan et al in 2012 & "corruption causes and cure" conducted by Haq (2012) were also theoretical studies.

The present study is different from previous studies as it is using time series data from the period (1984-2012) rather than cross- section data. It will empirically investigate the impact of economic determinants in Pakistan, existence of causal relation between corruption and its economic determinants in our country.

DESCRIPTION OF DATA

The study will use the annual data from world development indicators published by World Bank. The time series data is from 1984-2012 due to the availability of data on all variables. Income level will be proxied by GDP per capita growth (annual %), Consumer prices (annual %) will be used to measured as inflation rate, economic competition will be measured by degree of country’s openness proxied by exports of goods and services plus imports of goods and services as percentage of GDP, while FDI will be taken as percentage of GDP and population growth as annual percentage.

Data on corruption will be used from ICRG (International Country Risk Guide), an indicator of quality of government. It is the mean value of the ICRG variables "Corruption", "Law and Order" and "Bureaucracy Quality", scaled 0-1, where a higher value indicates higher quality of government i.e. less corruption and vice versa.

METHODOLOGY

To examine the impact of economic determinants of corruption following model will be estimated.

ICRG = α +α1 Y + α2 INFL + α3 FDI + α4 POP GW + α5 ECO COMP + µ

Where:

ICRG = International Country Risk Guide (ICRG)

Y = Income level proxied by GDP per capita growth (annual %)

INFL = Inflation rate proxied as consumer prices (annual %)

FDI = Foreign direct investment (% of GDP)

POP GW = Population growth (annual %)

ECO COMP = Economic competition measured by degree of country’s openness proxied by exports + imports as % of GDP

ARDL technique and Granger causality test will be applied respectively to check the impact of economic determinants of corruption and for the existence of causal relationship between corruption and its economic determinants in Pakistan.

ARDL technique is better to used for following reasons. Firstly, the variables considered can be integrated of order 1, 0 or both. Secondly, the model takes adequate number of lags to decrease the intensity of serial correlation of residuals in a general to specific modeling framework. Thirdly a dynamic error correction model (ECM) can be derived from ARDL through simple linear transformation (Shrestha and Chowdhury, 2005). Lastly, ARDL technique avoids the problem of endogeneity and gives unbiased estimates of the long-run model and valid t-statistics.