You are on page 1of 19

INTRODUCTION

Poverty is multidimensional. It is not only the lack of money but also the lack or deprivation of food, clothing,
shelter and other basic needs. This study aims to look at factors that could contribute to the either reduction or
increase of poverty. Argentina and Colombia are used for this study because the similarities in their culture and
availability of data on poverty.
THEORY
Foreign aid is inversely related to poverty.
In Thomas Malthuss view on population, poverty existed because of the increase on family size and population. In
other words, there is a direct relationship between population and poverty.
Unemployment is directly related to poverty
Mathematical Model
Phcr = a + b D1 - c fa_pop + d pop + e unemp +
Econometric Model
Phcr = 1 + 2country 3 fa_pop + 4 pop + 5 unemp +
Data
Let:
phcr = poverty headcount ratio at $1.25 (PPP) (% population) or percentage of population below $1.25 a day is
the percentage of the population living on less than $1.25 a day at 2005 international prices.
pop = population density or people per sq. km of land area
unemp = Unemployment refers to the share of the labor force that is without work but available for and
seeking employment. (expressed as percentage)
fa pop = total foreign aid received by country divided by the total population
Year

phcr

pop

unemp

country

fa pop

1991

0.55

12.09256

5.8

8.269882

1992

1.77

12.25569

6.7

7.867615

1993

2.05

12.4172

10.1

6.657635

1994

1.93

12.57737

12.1

4.248068

1995

3.88

12.73625

18.8

4.090069

1996

4.27

12.89408

17.2

3.706169

1997

3.77

13.05041

14.9

2.856229

1998

13.20378

12.8

2.187928

1999

16.18

35.24349

20.1

7.713287

1999

4.21

13.35227

14.1

2.095179

2000

17.85

35.83972

16.2

4.676069

2000

5.12

13.49466

15

1.42131

2001

19.18

36.43317

14.6

9.508543

2001

8.27

13.63038

18.3

3.902192

2002

20.26

37.02401

14.6

10.67234

2002

12.57

13.76018

17.9

2.164784

2003

19.61

37.61326

12

19.17885

2003

9.79

13.88584

16.1

2.806223

2004

18.98

38.20253

11.9

12.14442

2004

6.29

14.00991

12.6

2.38232

2005

12.71

38.79275

11.3

14.4171

2005

4.55

14.13429

10.6

2.486481

2006

11.04

39.38399

10.5

23.00503

2006

3.69

14.25951

10.1

2.948966

2007

8.84

39.97506

12

16.29723

2007

2.71

14.38529

8.5

2.574168

2008

11.32

40.56402

13.2

21.59745

2008

1.94

14.5118

7.8

3.287985

2009

9.67

41.1483

12

23.20605

2009

2.01

14.63902

8.6

3.16331

2010

8.16

41.72586

11.6

19.46394

Source: World Bank Database


Argentina : 1; 0 if otherwise

ECONOMETRIC ESTIMATION:
Dependent Variable: PHCR
Method: Least Squares
Date: 04/05/13 Time: 22:33
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
FA_POP
POP
UNEMP
C

-25.53890
-0.273859
-0.502158
0.365271
33.09468

10.56922
0.171569
0.456526
0.165326
16.72472

-2.416345
-1.596198
-1.099955
2.209397
1.978789

0.0230
0.1225
0.2814
0.0362
0.0585

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.817770
0.789735
2.839037
209.5635
-73.60821
29.16929
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

Phcr = 33.09 - 25.54country 0.27fa_pop 0.50pop + 0.36unemp

8.295806
6.191384
5.071497
5.302785
5.146891
0.693282

Interpretation of parameters:
1 : Holding all other factors constant, the average poverty headcount ratio for Colombia is 33.09468;
2 : Holding all other factors constant, the poverty headcount ratio in Argentina is 25.54 lower than Colombia or
7.55;
3 : Holding all other factors constant, a one dollar increase in the financial aid per capita will bring about a
0.273859 decrease in poverty headcount ratio;
4: Holding all other factors constant, a one unit increase in the population density will bring about a 0.502158 in
the poverty headcount ratio;
5: Holding all other factors constant, a one percent increase in the unemployment rate will bring about a
0.365271 increase in the poverty headcount ratio;
HYPOTHESIS TESTING
Confidence Interval: mean + z (se)
Coefficient Confidence Intervals
Date: 04/05/13 Time: 22:49
Sample: 1 31
Included observations: 31
95% CI
Variable

Coefficient

Low

High

COUNTRY
FA_POP
POP
UNEMP
C

-25.53890
-0.273859
-0.502158
0.365271
33.09468

-47.26425
-0.626524
-1.440561
0.025438
-1.283465

-3.813545
0.078807
0.436245
0.705103
67.47283

1 :

H0 : 1 = 0
HA : 1 0
= 0.05
95% confidence interval for 1 : - 1.283465 < 0 < 67.47283
Decision: within acceptance region therefore, accept null hypothesis

2 :

H0 : 2 = 0
HA : 2 0
= 0.05
95% confidence interval for 2 : -47.26425 < 0 < -3.813545
Decision: not within acceptance region therefore, reject null hypothesis
Therefore, there is a significant difference between the poverty headcount ratio in Argentina and
Colombia.

3 :

H0 : 3 = 0
HA : 3 0
= 0.05
95% confidence interval for 3 : -0.626524 < 0 < 0.078807
Decision: within acceptance region therefore, accept null hypothesis

4 :

H0 : 4 = 0
HA : 4 0

= 0.05
95% confidence interval for 4 : -1.440561 < 0 < 0.436245
Decision: within acceptance region therefore, accept null hypothesis
5 :

H0 : 5 = 0
HA : 5 0
= 0.05
95% confidence interval for 5 : 0.025438 < 0 < 0.705103
Decision: not within acceptance region therefore, reject null hypothesis

T-Statistics
1 :

H0 : 1 = 0
HA : 1 0
= 0.05
df = 26
|t-computed| > 2
1.978789 < 2
Decision: accept null hypothesis

2 :

H0 : 2 = 0
HA : 2 0
= 0.05
df = 26
|t-computed| > 2
|-2.416345 | > 2
Decision: reject null hypothesis
Therefore, there is a significant difference between the poverty headcount ratio in Argentina and
Colombia.

3 :

H0 : 3 = 0
HA : 3 0
= 0.05
df = 26
|t-computed| > 2
|-1.596198| < 2
Decision: accept null hypothesis

4 :

H0 : 4 = 0
HA : 4 0
= 0.05
df = 26
|t-computed| > 2
|-1.099955 | < 2
Decision: accept null hypothesis

5 :

H0 : 5 = 0
HA : 5 0
= 0.05
df = 26
|t-computed| > 2
2.209397 > 2
Decision: rejectt null hypothesis

1 :

0.0585 > 0.05


p value >
Decision: Accept H0

P-value

2:

0.0230 < 0.05


p value <
Decision: Reject H0

3:

0.1225 > 0.05


p value >
Decision: accept H0

4 :

0.2814 > 0.05


p value >
Decision: accept H0

5:

0.0362 < 0.05


p value <
Decision: Reject H0

F-statistics
H0: 2 = 3 = 4 = 5 = 0
H a: 2 3 4 5 0
29.16929 > 2.74
~together they affect the phcr
TEST FOR VIOLATIONS IN LINEAR REGRESSION MODEL:
A.) Test for multicollinearity

1.) correlation matrix


FA_POP
POP
UNEMP
FA_POP
1.000000
0.848575
-0.180014
POP
0.848575
1.000000
0.074711
UNEMP
-0.180014
0.074711
1.000000
There is high correlation between financial aid and population density.
2.) Auxiliary regressions
a.) Financial aid as possible suspect
Dependent Variable: FA_POP
Method: Least Squares
Date: 04/06/13 Time: 11:35
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

POP
UNEMP
C

0.486433
-0.479562
2.999198

0.049918
0.174282
2.515641

9.744693
-2.751635
1.192220

0.0000
0.0103
0.2432

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.779661
0.763923
3.405208
324.6723
-80.39395
49.53849
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.096671
7.008360
5.380255
5.519028
5.425491
1.069220

R-squared of 0.763923 is high. There could be possible presence multicollinearity.


VIF =
VIF =
VIF = 4.5384
VIF is within acceptable amount, there may not be a serious problem with multicollinearity

b.) Population density as possible suspect


Dependent Variable: POP
Method: Least Squares
Date: 04/06/13 Time: 12:42
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

FA_POP
UNEMP
C

1.587644
0.820773
-0.255633

0.162924
0.319195
4.658469

9.744693
2.571380
-0.054875

0.0000
0.0157
0.9566

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.773553
0.757378
6.151895
1059.683
-98.72903
47.82461
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

23.13667
12.48947
6.563163
6.701936
6.608400
0.879725

R-squared of 0.773553 is high. There could be possible presence multicollinearity


VIF =
VIF =
VIF = 4.41604
VIF is within acceptable amount, there may not be a serious problem with multicollinearity

Dependent Variable: UNEMP


Method: Least Squares
Date: 04/06/13 Time: 12:44
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

FA_POP
POP
C

-0.443849
0.232746
11.04743

0.161304
0.090514
1.340047

-2.751635
2.571380
8.244065

0.0103
0.0157
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.217247
0.161336
3.275962
300.4940
-79.19442
3.885580
0.032415

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

12.83871
3.577213
5.302866
5.441639
5.348103
0.666413

R-squared of 0.217247 is low. Presence of multicollinearity is unlikely.


VIF =
VIF =
VIF = 1.27754
VIF is very low, there may not be a serious problem with multicollinearity
B.) Test for Autocorrelation
1.) Plotting residual against actual and fitted data
25
20
15
10
6

2
0
-2
-4
-6
2

10

12

14

Residual

16

18

Actual

20

22

24

Fitted

26

28

30

There is a pattern or trend in the residual and actual data that could signify a possible presence of
autocorrelation.
2.) Scatter Diagram Residual over time

8
6

RESID

4
2
0
-2
-4
-6
1,988

1,992

1,996

2,000

2,004

2,008

2,012

YEAR

There is a funnel-like pattern in the scatter diagram of the residuals that could signify a possible
presence of autocorrelation
3.) Scatter Diagram lag one year

8
6

RESID(-1)

4
2
0
-2
-4
-6
-6

-4

-2

RESID

There is possible autocorrelation


4.) Durbin-watson:

The computed Durbin-Watson (d) is within the area of rejection. The null hypothesis being that there is no
positive/negative autocorrelation. Therefore, there is an evidence of positive autocorrelation.
C.) Test for Heteroscedasticity
White test:
Heteroskedasticity Test: White
F-statistic
Obs*R-squared

2.589793 Prob. F(13,17)


20.59882 Prob. Chi-Square(13)

0.0341
0.0813

Scaled explained SS

15.54282 Prob. Chi-Square(13)

0.2747

Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 04/06/13 Time: 13:54
Sample: 1 31
Included observations: 31
Collinear test regressors dropped from specification
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
COUNTRY
COUNTRY*FA_POP
COUNTRY*POP
COUNTRY*UNEMP
FA_POP
FA_POP^2
FA_POP*POP
FA_POP*UNEMP
POP
POP^2
POP*UNEMP
UNEMP
UNEMP^2

5413.494
-4724.017
-83.78442
209.0466
8.560592
126.3815
0.437886
-3.645820
0.070589
-327.8109
4.905131
0.383378
-10.97473
-0.014521

2648.631
2281.942
32.02742
87.84950
18.14012
47.93621
0.345080
1.565566
0.359776
138.5014
1.939248
0.693922
31.08336
0.227444

2.043884
-2.070174
-2.616021
2.379599
0.471915
2.636452
1.268939
-2.328756
0.196202
-2.366841
2.529398
0.552481
-0.353074
-0.063842

0.0568
0.0540
0.0181
0.0293
0.6430
0.0173
0.2216
0.0325
0.8468
0.0301
0.0216
0.5878
0.7284
0.9498

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.664478
0.407902
7.744937
1019.729
-98.13332
2.589793
0.034073

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

6.760112
10.06517
7.234408
7.882015
7.445511
1.446013

The computed chi-squared is 20.59882 is greater than the chi-squared on the table that is 15.379.
Therefore, there is a presence of heteroscedasticity.
REMEDIES:
Change functional forms:
1.) Log log
log phcr = 1 + 2country log 3fa_pop + log 4 pop + log 5 unemp +
Dependent Variable: LOG(PHCR)
Method: Least Squares
Date: 04/06/13 Time: 15:15
Sample: 1 31
Included observations: 31

Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
LOG(FA_POP)
LOG(POP)
LOG(UNEMP)
C

0.767321
-0.285859
2.282177
1.382523
-8.515398

1.277708
0.164693
1.137871
0.261666
4.475091

0.600545
-1.735707
2.005656
5.283548
-1.902844

0.5533
0.0945
0.0554
0.0000
0.0682

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.872063
0.852380
0.345981
3.112271
-8.358258
44.30623
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

1.785435
0.900492
0.861823
1.093111
0.937217
0.726203

P value for intercept, financial aid and population is higher than alpha of .05 and the sign of the
intercept does not meet my expectations in the model.
2.) Log lin
log phcr = 1 + 2country 3fa_pop + 4 pop + 5 unemp +
Dependent Variable: LOG(PHCR)
Method: Least Squares
Date: 04/06/13 Time: 15:18
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
FA_POP
POP
UNEMP
C

1.438562
-0.047700
0.130148
0.113722
-3.181288

1.475763
0.023956
0.063744
0.023084
2.335243

0.974792
-1.991138
2.041729
4.926408
-1.362294

0.3386
0.0571
0.0514
0.0000
0.1848

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.832049
0.806211
0.396410
4.085662
-12.57629
32.20187
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

3.) Lin log


phcr = 1 + 2country log 3fa_pop + log 4 pop + log 5 unemp +
Dependent Variable: PHCR
Method: Least Squares
Date: 04/06/13 Time: 15:17
Sample: 1 31
Included observations: 31

1.785435
0.900492
1.133954
1.365242
1.209348
0.825437

Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
LOG(FA_POP)
LOG(POP)
LOG(UNEMP)
C

-27.75136
-3.038738
-12.99972
3.908363
59.82216

10.95718
1.412353
9.757988
2.243955
38.37685

-2.532710
-2.151543
-1.332214
1.741729
1.558809

0.0177
0.0409
0.1944
0.0934
0.1311

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.800971
0.770352
2.967013
228.8823
-74.97502
26.15862
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.295806
6.191384
5.159678
5.390967
5.235073
0.554326

Drop a variable
1.) Drop fa_pop
phcr = 1 + 2country 3pop + 4 unemo +
Dependent Variable: PHCR
Method: Least Squares
Date: 04/06/13 Time: 15:24
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
POP
UNEMP
C

-32.21570
-0.897592
0.461770
42.87967

9.980601
0.394301
0.158225
16.00056

-3.227831
-2.276415
2.918438
2.679885

0.0033
0.0310
0.0070
0.0124

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.799913
0.777681
2.919281
230.0995
-75.05722
35.98043
0.000000

2.) Drop pop


phcr = 1 + 2country 3fa_pop + 4 unemo +
Dependent Variable: PHCR
Method: Least Squares
Date: 04/06/13 Time: 15:24

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.295806
6.191384
5.100466
5.285497
5.160781
0.718579

Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
FA_POP
UNEMP
C

-14.12734
-0.376267
0.374605
15.19157

2.026946
0.144670
0.165748
3.862073

-6.969765
-2.600865
2.260079
3.933527

0.0000
0.0149
0.0321
0.0005

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.809290
0.788100
2.850052
219.3154
-74.31321
38.19218
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.295806
6.191384
5.052465
5.237496
5.112781
0.528258

3.) Drop unemp


phcr = 1 + 2country 3fa_pop + 4 pop +

Dependent Variable: PHCR


Method: Least Squares
Date: 04/06/13 Time: 15:25
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
FA_POP
POP
C

-28.72776
-0.412473
-0.553931
42.05891

11.19753
0.170780
0.487595
17.35220

-2.565544
-2.415239
-1.136046
2.423837

0.0162
0.0228
0.2659
0.0223

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.783557
0.759508
3.036253
248.9085
-76.27512
32.58143
0.000000

4.) Drop fa_pop and pop


phcr = 1 + 2country + 3 unemp +

Dependent Variable: PHCR


Method: Least Squares
Date: 04/06/13 Time: 15:27

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.295806
6.191384
5.179040
5.364070
5.239355
0.509843

Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
UNEMP
C

-9.629917
0.576842
6.792110

1.161298
0.160740
2.325851

-8.292372
3.588655
2.920269

0.0000
0.0013
0.0068

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.761511
0.744476
3.129708
274.2621
-77.77860
44.70282
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.295806
6.191384
5.211522
5.350295
5.256759
1.024152

FINAL MODEL:
Phcr = 1 + 2country 3fa_pop + 4unemp
Dependent Variable: PHCR
Method: Least Squares
Date: 04/06/13 Time: 15:24
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

COUNTRY
FA_POP
UNEMP
C

-14.12734
-0.376267
0.374605
15.19157

2.026946
0.144670
0.165748
3.862073

-6.969765
-2.600865
2.260079
3.933527

0.0000
0.0149
0.0321
0.0005

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.809290
0.788100
2.850052
219.3154
-74.31321
38.19218
0.000000

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.295806
6.191384
5.052465
5.237496
5.112781
0.528258

Phcr = 15.19 14.13country 0.38fa_pop + 0.37unemp


Interpretation of parameters:
1 : Holding all other factors constant, the average poverty headcount ratio for Colombia is 15.19157;
2 : Holding all other factors constant, the poverty headcount ratio in Argentina is 14.12734 lower than
Colombia;
3 : Holding all other factors constant, a one dollar increase in the financial aid per capita will bring about
a -0.376267decrease in poverty headcount ratio;

4: Holding all other factors constant, a one percent increase in the unemployment rate will bring about a
0.374605 increase in the poverty headcount ratio;
TEST FOR VIOLATIONS ON LINEAR REGRESSION MODEL
A.) Test for multicollinearity:
1.) Correlation Matrix:
FA_POP
1.000000
-0.180014

FA_POP
UNEMP

UNEMP
-0.180014
1.000000

Very low correlation; low possibility of presence of multicollinearity


2.) Auxillairy Regressions and Variance Inflation Factor
Dependent Variable: FA_POP
Method: Least Squares
Date: 04/07/13 Time: 00:35
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

UNEMP
C

-0.352678
12.62460

0.357865
4.763992

-0.985505
2.650004

0.3325
0.0129

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.032405
-0.000960
7.011724
1425.764
-103.3285
0.971220
0.332528

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

8.096671
7.008360
6.795385
6.887901
6.825543
3.158815

VIF =
VIF =
VIF = 1.033
There is no presence of multicollinearity

Dependent Variable: UNEMP


Method: Least Squares
Date: 04/07/13 Time: 00:41
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

FA_POP
C

-0.091883
13.58266

0.093234
0.991484

-0.985505
13.69931

0.3325
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.032405
-0.000960
3.578930
371.4534
-82.48036
0.971220
0.332528

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

12.83871
3.577213
5.450346
5.542861
5.480503
1.254813

VIF =
VIF =
VIF = 1.033
There is no presence of multicollinearity
B.) Test for Autocorrelation:
1.) Scatter Diagram Lag one year
8
6
4

RESID(-1)

2
0
-2
-4
-6
-8
-8

-6

-4

-2

RESID

The scatter diagram does not show any pattern that could determine a presence of
autocorreation

C.) Test for Heteroscedasticity:

Heteroskedasticity Test: White


F-statistic
Obs*R-squared
Scaled explained SS

3.098444 Prob. F(6,24)


13.53138 Prob. Chi-Square(6)
7.527443 Prob. Chi-Square(6)

0.0216
0.0353
0.2748

Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 04/07/13 Time: 00:43
Sample: 1 31
Included observations: 31
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
COUNTRY^2
COUNTRY*FA_POP
COUNTRY*UNEMP
FA_POP^2
FA_POP*UNEMP
UNEMP^2

17.09951
-45.33370
0.932483
3.118968
-0.141107
0.304853
-0.105017

18.51716
29.83057
1.790135
1.545324
0.062238
0.161792
0.056577

0.923441
-1.519706
0.520901
2.018326
-2.267211
1.884221
-1.856193

0.3650
0.1417
0.6072
0.0549
0.0327
0.0717
0.0757

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

0.436496
0.295620
11.08786
2950.578
-114.6015
3.098444
0.021629

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

11.12892
13.21127
7.845259
8.169063
7.950811
2.206336

The computed chi-square, 13.53138 is less than the chi-square distribution of 16.1513.
Therefore, there is no heteroscedasticity
HYPOTHESIS TESTING:
A.) Confidence intervals:
Coefficient Confidence Intervals
Date: 04/07/13 Time: 00:49
Sample: 1 31
Included observations: 31
95% CI

Variable

Coefficient

Low

High

COUNTRY
FA_POP
UNEMP
C

-14.12734
-0.376267
0.374605
15.19157

-18.28629
-0.673105
0.034517
7.267250

-9.968390
-0.079429
0.714693
23.11589

H0 : 1 = 0
HA : 1 0
= 0.05
95% confidence interval for 1 : 7.267250 < 1 < 23.11589
Decision: not within acceptance region therefore, reject null hypothesis

2:

H0 : 2 = 0
HA : 2 0
= 0.05
95% confidence interval for 2 : -18.28629 < 2 < -9.968390
Decision: not within acceptance region therefore, reject null hypothesis

3:

H0 : 3 = 0
HA : 3 0
= 0.05
95% confidence interval for 3 : -0.673105 < 3 < -0.079429
Decision: not within acceptance region therefore, reject null hypothesis

4:

H0 : 4 = 0
HA : 4 0
= 0.05
95% confidence interval for 4 : 0.034517 < 4 < 0.714693
Decision: not within acceptance region therefore, reject null hypothesis

B.) T-statistics
1 :

H0 : 1 = 0
HA : 1 0
= 0.05
|t-computed| > 2
3.933527 > 2
Decision: reject H0

2:

H0 : 1 = 0
HA : 1 0
= 0.05
|t-computed| > 2
|-6.969765| > 2
Decision: reject H0

3:

H0 : 3 = 0
HA : 3 0
= 0.05
|t-computed| > 2
| -2.600865 | > 2
Decision: reject H0

4:

H0 : 4 = 0
HA : 4 0
= 0.05
|t-computed| > 2
2.260079 > 2
Decision: reject H0

C.) P-value
1 :

0.0005 < 0.05


p value <

Decision: Reject H0
2:

0.0000 < 0.05


p value <
Decision: Reject H0

3:

0.0149 < 0.05


p value <
Decision: Reject H0

4:

0.0321 < 0.05


p value <
Decision: Reject H0

D.) F-statistics
H0: 2 = 3 = 4 = 0
H a: 2 3 4 0
3.098444 > 2.95
~together they affect the phcr

USE THE MODEL FOR POLICY


Since there is a significant difference in the poverty headcount ratio between Colombia and Argentina,
Colombia could look into the poverty alleviation programs of Argentina and apply the similar policy that will fit into
their situation. The unemployment rate has a significant impact in poverty for both countries. Both should look at
creating jobs for the people in order to unemployment and in turn, reduce poverty. The foreign aid also has an
effect in the poverty situation of both countries. International organizations could increase the aid or assistance to
countries with high poverty rates and it should also check into the increasing the effectiveness of the aid given to
other countries.

You might also like