Regression analysis: Simple Linear Regression Multiple Linear Regression
Farhan research
1. 1
Chapter 1
(1.1) Introduction
FDI (Foreign direct investment) is one of the most important aspects to see the rate of
development of a country. We also know that FDI is used as a way of bringing the developing
countries in to the global market place and also serves as a great source of eradicating poverty in
these developing countries. FDI in Pakistan was about 800 million US dollars in 2012 which
was down from 1.3 billion US dollars in 2011. As agriculture is one of the major sectors in the
economy of Pakistan. It is greatly affected by these fluctuations.
In this paper the most important thing is to study the impact of FDI on agriculture
employment level. Governments of different countries go to big lengths to attract foreign direct
investment. The positive effect of foreign direct investment on the economy as a whole is
substantial. Some of these expected returns are that an increase in FDI may cause an increase in
capital in local market through various channels, it may also cause an eventual technology spill-
over which will in turn cause an increase an increase in the development of domestic firms.
FDI is particularly important for developing countries which get access to resources which
they would not be able to get without any FDI. For only this one purpose they go through great
measures to attract foreign direct investment.
The agriculture sector is also affected largely by the FDI. As the FDI increases. The
Consumption of people also increases when eventually new industries are constructed. The raw
material they need is provided by the agriculture sector. So as the FDI increases and the demand
for consumable goods and Raw materials increases. The employment in Agriculture also
increases to meet the demand.
This study seeks to answer just one basic question. That is to find what the amount of change
in agriculture employment level is due to change in FDI. For this purpose time series data was
used that consisted of two variables and years ranging from 1980 to 2009.
(1.2)Specification of the Problem
2. 2
The question we want to answer in this paper is to find out what amount of change in
employment level in agriculture sector is caused by Foreign Direct investment and what amount
of change is caused by other variables.
(1.2)Research Purpose
The main purpose of this research is to find out that what amount of change in agriculture
employment level is caused by foreign direct investment.
(1.3)Objective
There are many variables which causes changes in the employment level in agriculture
sector of Pakistan. The sole purpose of my research is to find out how much of this change is
caused due to change in FDI.
(1.4)Hypothesis
H0: There is no significant relation between FDI and the employment level in agriculture sector.
H1: There is a significant relation between FDI and the employment level in agriculture sector.
Organization of the Study
3. 3
Introduction
Objective
Hypothesis
Literature Review
Theoretical Framework
Research Methodology
Descriptive analyses of the Data and Econometrics Tests
Summary and conclusion
References
Chapter 2
Literature Review
The neo classical theory states that FDI can increase the capital available per person. FDI
also causes long term growth by influencing many variables e.g. research and development and
human capital. The FDI also causes spill over principle affect to unrelated firms in the host
economy. (Mustapha Hussein 2003)
According to Bloomstorm and Koko (2003) and Bornstein, De gregerio and lee(1998), FDI is
widely contributed with filling the Gap between the domestically available resources and the
desired investment. It is also contributed with improving the labor and management skill of the
4. 4
host country. These skills can help the country to get out of the poverty. (Bornstein, De gregerio
and lee 1998).
Many developing countries in the early nineties put all their strife towards attracting the
foreign direct investment. They introduced new trade and political measures that could enhance
and propagate suitable setting for the growth of foreign direct investment. Most notable reforms
introduced were the liberalization policy along with entering into various multilateral accords
that promote confidence among the foreign investors. Between 2005 to 2008 the increase in
foreign direct investment in Africa was reported around 80 percent, despite global financial
crises (2008). The foreign direct investment has doubled from 2001 to 2007 in the agribusiness
(Weissleder, L., 2009).
Foreign direct investors not just bring in investment inflows but also provide various
opportunities like business knowledge, approach to international market and new technologies to
the host country. Agricultural growth and development are paramount to healthy economy and
poverty reduction in developing nations as economies of such countries are heavily dependent on
agricultural sector. These parameters like new technologies and knowledge make way for the
growth of agricultural sector in the host country. Both agricultural and developmental
economists agree that agricultural output can be increased by growth and development in
agricultural sector (Namizinga N. 2007).
The technological changes along with other effective reforms magnify the export earnings of
the country that help in maintaining healthy foreign reserve. With improved foreign reserve
capacity as a result of enhancement in agricultural growth, the developing country is able to
import the up to date technology that is beneficial in the overall transformation of the agricultural
sector. Revenues from agricultural sector in Ethiopia provides its government with the proceeds
to import better quality seed and fertilizer to improve growth in the agriculture (Bijsterbosch, M.
and M. Kolasa. 2009).
5. 5
Elibarik M. (2007) is one of the very few researchers who have studied the impact of foreign
direct investment on agriculture in the developing countries. In his research he stated that cash
inflows from foreign direct investors into the agricultural sector are vital for several main
reasons. Agricultural sector has a reasonable role in the economy where it increases growth and
help reduces poverty. The strategies and policies designed to reduce poverty in countries where
the majority of population lives on agriculture must be formulated on the basis of agricultural
growth and development. Foreign direct investment enhances employment rate which in turn
increases the per capita income and participate to large extent in the reduction of poverty.
Karlsson et al, Nannan lundin, Fredrick Sjoholm and Ping He (2007) investigated the impact
of foreign direct investment on creation of jobs in China. The findings materialized that foreign
direct investment and private firms created more jobs than public or non private enterprises. The
examination on the basis of ownership declared that foreign direct investment firms were in a
better and more advantageous position than others regarding employment growth. Axarloglou
and Pournarakis (2007) found out the impact of foreign direct investment on local employment
in non agricultural sector. The outcome showed varying effects from industry to industry. The
foreign direct investment had an affirmative impact on job creation in industries like printing and
publishing, transportation and instruments and equipments while negative impact on
employment in leather, clay and glass industries.
Wang and Zhang (2005) using the data from 1984- 2004 probed that both the productivity
levels and foreign direct investment in combination had the positive impact on employment in
Chinese economy. Many determinants contribute to foreign direct investment inflows. Exchange
rate could be one of the factors. Cheng and Yu measured the impact of RMB exchange rate on
foreign direct investment. Russ (2007) found out that the decisions of MNCs entering a certain
market could depend on the exchange rate fluctuations and might affect their decisions.
6. 6
Chapter 3
Theoretical Framework
(4.1)Variables
Dependent: AG (Agriculture sector employment level)
Independent: FDI (Foreign Direct Investment)
AG stands for agriculture sector employment level. By this we mean anyone who is
employed in the agriculture sector. This include working on one’s own farm, dairy farms or
someone else farms. This also includes people whose jobs are related to this sector i.e. the
middlemen who are involved in transporting the goods from fields to industries and markets.
By FDI we mean foreign direct investment which means the net sum of all the investment
that comes to Pakistan. The level of FDI greatly effects the development of various sectors and
agriculture is one of them. And we want to find out what amount of change can FDI cause to
employment level in the particular sector of agriculture.
AG is going to be our dependent variable because we want to find out the effect of FDI on
AG. That’s the reason FDI will be our independent variable.
(4.2)How does FDI actually affect employment level in agriculture sector?
According to me and the literature review done by for me by this research i.e.(Ikira 2003).
There are two main channels through which FDI affects agriculture sector. Firstly when FDI in a
country increases, the amount of capital in that particular nation also increases i.e. people have
more money to spend. Hence demand for goods both consumable and raw materials increases.
To meet this demand the agriculture sector need to increase its production which means they
have to increase their inputs. And one of the important inputs is Labour. In this way more labour
is employed.
Second channel is when FDI increases and new industries are constructed and/or the
productive capacity of the present industries is increased. The raw material that is demanded
7. 7
from the agriculture sector increases and to meet this demand the inputs in the agriculture sector
are also increased. And in this was the employment level also increases.
The same two channels can be used to show what will happen if the FDI in a country
decreases.
Chapter 4
Research Methodology
(5.1)DATA
Secondary data was used which was taken from the World Bank site.
(5.2)Sample
30 observations were used with years ranging from 1980 to 2009.
(5.3)Model
The Data was analyzed using Gretl. The data was first checked for stationarity using
augmented dickey fuller test. Both the variables were non stationary. At first difference both the
variables became stationary.
8. 8
Then a simple OLS model was run because we had just one independent variable and one
dependent variable and we just wanted to analyse the R-square value to answer our research
question that how much amount of change in Agriculture level employment is due to FDI.
Simple linear regression model was used.
OLS model : Independent variable: FD
Dependent variable : AGI
Chapter 5
Descriptive analyses of the Data and Econometrics Tests
(5.1)Data Table
years FDI AG
1980 0.268611 51.10746
1981 0.384635 47.05429
1982 0.207752 48.38725
1983 0.102667 47.06574
1984 0.178192 50.18642
1985 0.421864 46.89026
1986 0.331453 46.80542
1987 0.387921 44.21438
1988 0.484737 47.09245
1989 0.387921 47.32854
1990 0.612998 48.46821
1991 0.641556 48.49815
1992 0.691844 42.19854
1993 0.677094 42.10984
1994 0.811304 43.09546
1995 1.191752 43.40986
1996 1.456054 43.69432
1997 1.147229 43.65982
1998 0.863333 44.79457
1999 0.844795 42.42982
9. 9
2000 0.416484 42.06544
2001 0.529666 41.70105
2002 1.138205 41.33667
2003 0.641482 40.97228
2004 1.141075 40.60789
2005 2.008212 25.68483
2006 3.351373 29.78543
2007 3.904417 34.87372
2008 3.318045 39.15035
2009 1.444824 20.54078
(5.2)Augmented Dickey fuller test for both variables to check if Data is
stationary
FDI
Augmented Dickey-Fuller test for FDI
including 8 lags of (1-L)FDI (max was 8)
sample size 21
unit-root null hypothesis: a = 1
test with constant
model: (1-L)y = b0 + (a-1)*y(-1) + ... + e
1st-order autocorrelation coeff. for e: -0.287
lagged differences: F(8, 11) = 4.150 [0.0163]
estimated value of (a - 1): 0.0647967
test statistic: tau_c(1) = 0.186162
asymptotic p-value 0.9718
The variable FDI is not stationary in test with constant as the p value is more than 5%. Now we
will check if it is stationary at first order difference.
Augmented Dickey-Fuller test for d_FDI
including 8 lags of (1-L)d_FDI (max was 8)
sample size 20
unit-root null hypothesis: a = 1
test with constant
model: (1-L)y = b0 + (a-1)*y(-1) + ... + e
1st-order autocorrelation coeff. for e: -0.208
lagged differences: F(8, 10) = 5.029 [0.0102]
estimated value of (a - 1): -9.03613
test statistic: tau_c(1) = -4.30364
asymptotic p-value 0.0004328
10. 10
The variable FDI is stationary at first order difference as p value is less than 5%.
AG
Augmented Dickey-Fuller test for AG
including 3 lags of (1-L)AG (max was 8)
sample size 26
unit-root null hypothesis: a = 1
test with constant
model: (1-L)y = b0 + (a-1)*y(-1) + ... + e
1st-order autocorrelation coeff. for e: -0.032
lagged differences: F(3, 21) = 9.226 [0.0004]
estimated value of (a - 1): 0.3574
test statistic: tau_c(1) = 2.12328
asymptotic p-value 0.9999
The variable AG is not stationary. Now we will check if it becomes stationary at first order
difference.
Augmented Dickey-Fuller test for d_AG
including 2 lags of (1-L)d_AG (max was 8)
sample size 26
unit-root null hypothesis: a = 1
test with constant
model: (1-L)y = b0 + (a-1)*y(-1) + ... + e
1st-order autocorrelation coeff. for e: 0.190
lagged differences: F(2, 22) = 9.321 [0.0012]
estimated value of (a - 1): -3.04506
test statistic: tau_c(1) = -6.78439
asymptotic p-value 1.247e-009
The variable AG became stationary at first order difference as the value is way less than 5%.
Both the variables FDI and AG becomes stationary at first order difference
Now we will run the OLS regression with FDI as the independent variable and AG as dependent
variable.
OLS model
11. 11
Model 1: OLS, using observations 1980-2009 (T = 30)
Dependent variable: AG
Coefficient Std. Error t-ratio p-value
Const 47.1672 1.4139 33.3596 <0.00001 ***
FDI -4.66215 1.02668 -4.5410 0.00010 ***
Mean dependent var 42.50697 S.D. dependent var 6.897595
Sum squared resid 794.5697 S.E. of regression 5.327053
R-squared 0.424111 Adjusted R-squared 0.403544
F(1, 28) 20.62049 P-value(F) 0.000097
Log-likelihood -91.71721 Akaike criterion 187.4344
Schwarz criterion 190.2368 Hannan-Quinn 188.3309
Rho 0.000906 Durbin-Watson 1.467241
Now we will run White test to check whether the Data is Hetroskedastic or
Homoskedastic
White's test for heteroskedasticity
OLS, using observations 1980-2009 (T = 30)
Dependent variable: uhat^2
coefficient std. error t-ratio p-value
-------------------------------------------------------
const −37.9847 31.4452 −1.208 0.2375
FDI 111.746 52.6427 2.123 0.0431 **
sq_FDI −24.9028 13.5427 −1.839 0.0770 *
Unadjusted R-squared = 0.160313
Test statistic: TR^2 = 4.809377,
with p-value = P(Chi-square(2) > 4.809377) = 0.090294
As the p value is more than 5%. We accept the null hypothesis. Which means that the data is
homoscedastic.
Analysis of tests
The ADF test shows that the Data is stationary with constant on first order difference.
12. 12
The White test shows that the Data is Homoskedastic and the variables don’t depend on their lag
values.
DW statistic is near to 2 i.e. 1.46 . This indicates that the data is tending towards normality and
the auto correlation is not that severe and we can use the OLS model for our study.
The regression has an R-squared value of 0.424111. This means that 42.4% of change in AG is
due FDI and the other 57.6% is due to other variables.
In layman’s terms about 42.5 % of change in agriculture employment is due toForeign Direct
investment.
Chapter 6
(6.1) Summary and Conclusions
This research first used different tests to prove that the data is not biased. And then simple
OLS model was used to find out that 42.4% of change in Agriculture employment level is due to
the foreign direct investment. This is a very substantial amount of dependency. And using this
finding various steps can be taken to increase the FDI which will in turn increase the welfare of a
large sector of people in Pakistan
(6.2) Policy Recommendations
The recommendation I can give based on this study I conducted is that FDI is one of the
major factors in determining the employment level in the agriculture sector. And as Pakistan is
still mostly agriculture based economy, any steps taken to increase employment in this sector
will increase the overall welfare of Pakistan. For this reason steps should be taken to attract
foreign direct investment in to the nation. This incentives can either be economic or some other
type of other incentives. But if we want development in Pakistan’s economy. We need to
increase our FDI.