Economical determinants of domestic

5.Data and empirical results
5.1.Sources of Data
The annual data for Jordan was computed from different sources and the time period
spanning from 1980 to 2005. The data definitions and statistical sources used in this study are
listed in Table (2).
Table 2: Description of the Data and Statistical Sources
Variable Description of the Data Source
Gr Growth rate of real GDP CBJ, DOS
FDI Foreign Direct Investment as a ratio of GDP CBJ, DOS
GDI Gross Domestic Investment (net of FDI)
as a ratio of GDP
CBJ, DOS
FI Financial intermediation proxied by M2 /GDP CBJ
H Human capital proxied by secondary school
enrolment ratio
UNESCO, DOS, MOE
X Export of goods and services
as a ratio of GDP CBJ, DOS
Cr Domestic credit
CBJ, DOS
Note: CBJ, denotes Central Bank of Jordan; DOS, Departments of Statistics,
International finance Statistics, MOE, Ministry of Education.
5.2 Long--Run Equilibrium Estimation
To explore the existence of long-run relationship or co-integration among the variables
in the model (7), we employed the ARDL test proposed by Pesaran et al., (2001). The
existence of a unique long-run relationship is crucial for valid estimation and inference about
the parameters of the models. One of the important advantages of the ARDL procedure was
that the estimation is possible even when the explanatory variables are endogenous

---

European Scientific Journal
April edition vol. 8, No.7
ISSN: 1857 – 7881 (Print)
e -
ISSN 1857- 7431
9
(AlamandQuazi, 2003). Furthermore, as long as the ARDL model is free of residual
correlation, endogeneity is less of a problem. Pesaran and Shin (1999) showed that the
appropriate lags in the ARDL model are corrected for both residual correlation and
endogeneity.
The Schwarz Bayesian Criteria (SBC) selects an ARDL (1,0,0,1,0,0) domestic
investment model for Jordan. The results of ARDL model of Jordan's domestic investment are
reported in Table (3). As it can be seen from the table, most of the variables are significant
and the signs are consistent with a priori expectation. With the exception of credit availability,
real GDP growth, FDI, financial intermediation and exports statistically have a significant
positive effect on domestic investment.
The long-run results of ARDL for Jordan’s data indicate that the expansion in the
export of goods and services is a key determinant of domestic investment. The coefficient of
exports indicated that exports had a larger impact on domestic investment, which suggest that
a 1 million JD increase in the exports could result in 62 million JD increase in domestic
investment accumulation. Our result confirms similar results of Jansen, 1995, and Cuvers
(1996) for groups of developing countries. In addition, the estimation results provided support
for a complementarity between domestic investment and foreign direct investment (FDI) that
means FDI has a strong stimulus effect on domestic investment. This suggests that a 1 million
JD increase in FDI could result in an increase in domestic investment by 17 million JD, which
indicated the inflow of FDI “crowds in” domestic investment in Jordan, and confirming a
similar finding of Borensztein, et al., (1998).
The long-run estimation indicates that improvement of financial intermediation
(captured by ratio of browed money to GDP) is boosting domestic investment by their
contribution to lowering the requirement to finance and thereafter lowering the cost of
borrowing. This result is in line with the hypothesis that financial intermediaries provide the
link between the financial and the real sector and confirmed theoretical literature arising out
of the McKinnon and Shaw hypothesis (McKinnon 1973, Shaw 1973) and confirmed the
findings of Fry (1998); Ghura and Goodwin (2000), and Agrawal (2000).
As shown on the table, the real growth of real GDP is an important determinant of
domestic investment, confirming the results of Greene and Villanueva (1991) and Ghura and
Goodwin (2000), and it is consistent with the findings of Blomstrom et al. (1996) where there
is a causality between economic growth and gross investment and confirmed the result of
Madsen, (2002) regarding policies that seek to enhance investment which are the effective
means of promoting economic growth. Booth (1999) argued that rapid growth leads to high

---

European Scientific Journal
April edition vol. 8, No.7
ISSN: 1857 – 7881 (Print)
e -
ISSN 1857- 7431
10
rates of investment and vice versa. See also: De Long and Summers, (1992); Chaudhari and
Wilson, (2000); Podrecca and Carmeci, (2001), and Krishnaa et al., (2003).
Table 3. Long-Run Coefficients Estimates for Economic Growth
Model
Regressors Dependent Variable Domestic Investment
SBC Selected model
(1,0,0,1,0,0)
Coefficients T-ratio [P-value]
Gr 0.4474 2.6446 [.018]**
X 0.6201 2.4405 [.028]**
FI 0.6188
2.2908 [.037]**
FDI 0.1721 3.2684 [.005]***
H 0.0994 2.5846 [.021]*
CR
0.1243 [0.7395 [.825]
C
-2.3256
-1.9060 [.076]*
The period [No. of Obs.] 1980-2010 [36]
Note: Following Pesaran et al. (2001), lag order of the ARDL model was selected
using Schwarz Bayesian Criteria (SBC) and the LM tests for testing residual correlation.
Asterisks ***, **, * represent 1%, 5%, 10% significant levels, respectively. The t-ratios are
reported in square brackets. The following notation applies: Gr, denotes growth rate of gross
domestic product; GDI, gross domestic investment; FDI, foreign direct investment; FI,
financial intermediation (M2/GDP); H: human capital; X: exports of goods and services. T:
not significant thus omitted from the table.
Another interesting result arising from long-run ARDL relates to the important role of
educational development in Jordan in stimulating domestic investment which is consistent
with theories saying that the higher the educational development (as proxy for the human
capital) the higher the level of domestic investment (see: Borensztein et al., 1998 and Ghura
and Goodwin, 2000). While the increase in domestic credit did not appear to have a
significant impact on domestic investment in the long-run, but it had a significant impact in
the short-run as will be seen in the ECM result in the next section.

---

European Scientific Journal
April edition vol. 8, No.7
ISSN: 1857 – 7881 (Print)
e -
ISSN 1857- 7431
11