feder (1983)

Prévia do material em texto

JtllLrnal of Development Economics f 2 (1982) 59-73. North-Holland Publishing Company 
ON EXPORTS AND ECONOMIC GROWTH* 
Gershon FEDER 
World Bank, Washington, DC 20433, USA 
Received October 1981, final version received March 1982 
The popper analyses the sources of growth in the period 1964-1973 for a group of semi- 
industrialized less developed countries. An analytical framework is developed, incorporating the 
possibility that marginal factor productivities are not equal in the export and non-export sectors 
df the economy. Econometric analysis utilizing this framework indicates that margmal factor 
productivities are significantly higher in the export sector. The difference -VI, Michalopoulos and Jay ( 19’73), Tyler ( 198 I)]. 
Explanations for these observations were discussed by a number of 
economists, highlighting various beneficial aspects of exports, such as greater 
ca‘pacity utilizaiion, economies of scale, incentives for technological 
improvements and effkient management due to competitive pressures abroad 
[see 3alassa (19’78), Meesing (1967, 1’269) Krueger ( 1980), and Bhagwati and 
Sriniva.san ( 1978)]. The implication of ese discussions is that there a-e 
substaCa1 differences between margin&l factor productivities in export- 
orlented and non-export-oriented industries, such that the former, have Sgher 
“The views expred in this pa r are those cd the author only and cannor be taken 10 
represent the views of the World Bank. The authclr e&d from helpful discu:.sion3 with R. 
Sk?, Keesing, 6. Bcmell. S. Robinsor.. M. Syrquin 
and M. Wolf. 
03(~3878:82/~~/$02.75 0 1382 Xorth-Holland 
ADEILSON
Highlight
factor productivity, It fallows, then, that counuties which have adopted 
policies less biased against exports benefited from closer&-optimal resource I” 
allocation and higher growth, 
The present paper develops an anrilytical framework for tbe quantitative 
assessment of f&or productivity di@ere&ls Uxtqween expc>rts and non- 
expo~s using ~~~~~~~ ,‘Pd”, ,:“cfii~’ ~f~~~~~i i$ aii[$& i$’ in ern~iirical 
study of so~rrzes of growth in a sample of, Islemi-industrialized less developed 
countries within the decade S@Pi9?3. The ‘analysis allows gm estimate of 
the sectoral bargina productivities. FurthermoreY the extent of inter-sectoral 
beneficial exteqnalities generated by exports can be specifically ideniitied. The 
results highlight the role of eliport performahce in explaining the groT#th 
record of the sample counties. 
m a-nalytis in this paper adope a sup& side description of changes in 
aggregate ou:put. In SO doing it follows a practice widely used h the 
empirical study of sources of growth [see, e.g., Hagen and I-Iawrylyshyn 
(;t969) and Robinson (197 l)]. Within such a framework, where aggregate 
growth is ;related to changes in capital and labor through an underlying 
production function, the ‘.’ stub:es of Balassa (197P), Chenery et al. (1970), 
Michalopoulos and Jay (1973) and Tyler (1981) havl? included an indicator of 
export performance among the variabies explaining growth.” In the 
following, a framework is developed which provides a formal rationalization 
for t5e incorporation of export varizlzles in the sources of growth equation. 
Fptihenno& starting from the sectoral production functions, a proper 
smcatioti of export v&iables is indicated and a non-conventional 
intizrpretation of parameters is implied. 
Since the present analysis focuses on the potential non-optimaiity of 
resource allocation between export and non-export sectors, the economy is 
Gtwe(‘t as if it cclnsists of two ‘distil- ct sectors: one producing export goods, 
he other producing for the &. lnestlc market.2 Instead of an aggregate 
al productioln function, eacl. J 1 the tvro sectors” output is a function Qf 
the falztors alloca@:d to the sector, En addition, the output of the non-export 
sector is dependent OTI Ilhe volume of exports produced. This formulatisn 
represents the beneficial effects of exports 011 other sectors [Keesing (1967, p. 
511, 14’79, pp. 4, !5)& such as the development of eflicient and internationally 
rqresent exports pzformarrce by export g”cr\wth multiplied by export share; 
e expsrts _growtIiL, 
n s.bstraction, as many firms are pxxlucing for both domestic and external 
argued that even then, the domestically marketed output su& firms is 
e a’ame fact:tc*r prQd~~~ti~~ti~s which characterize exports. wever, to the 
exports represents a good approximation of the changes in the \ olume of 
~~~~~~~t~~~ of such firms, the results are stilt valid. 
ADEILSON
Highlight
ADEILSON
Highlight
compe; itive mana lent, the introduction of improved production 
her quality labr, steadier flow of imported inputs, 
r-red to tos externalities, since they aire not reflected irr 
are incorporated n the formulation below: 
N = F(K,, L,, X), (1) 
X = G(K., Lx), where (21 
N = non-exports, 
X = exports, 
K,, K, = respective sector caprtal stocks, 
L Lx = respective sector labor forces. 
Since data regarding sectcsal allocations of primaisy production factors are 
no! readily available, a speScation is required which will allow estimates of 
sectoral marginal productivities using aggregate (national) data. This is 
accomplished as follows:. 
Suppose that the ratio of respective marginal factor productivities in the 
two sectors deviates from unity by a factor 6, i.e., 
(G,/Fk)=(G,/FL)= I+& 13 
where the subscripts denote partial derivati v es. 
In the absence of exiernalities, and for a given set of prices, a situation 
where 6 =O would reflect an allocation of resources which maximizes 
national output. However, due to a number of reasons, marginal factor 
produe&ities are likely to be lower in the non-export sector (i.e., 6 ~0). One 
important reason is the more competitive environment in which export- 
oriented firms operate. Competition induces innovativeness, adaptabilny, 
efficient management of firms‘ resources, etc. Another reason for deviations 
between sectoral margi ctor productivities are various regulations and 
constraints such as cre x&ange rationing [Balbssa ( I977)3. 
e higher perceived u iatecl with export enterprises may be 
anotner factor contrib mdaginai 
Productivit y diflfereh Gals to externalities arc not i 
as y will b;; specificall 
ifferer Liation of eq 
where Z,, and 1, ;are respective sectoral gross investzlents, t, and LX are 
see&al- zbanga in labor - ha, and- F, describes. the lnarjnal externality 
43&&t of e3QW%3 4Bk %Woutprut of non*xprts. ’ *I : ’ 
Denoting Gross Domer;tic Product by Zy; and since by definrtion Y=4V -o- 
X, it follows Y 
2 
k=N+d. (6) 
Using eqs. (.3)-(5) in eq. (6) yie’ids 
Define total investment I (~2, + I,J and total growth of labor i. (E& + 
f,). Rec&‘that eqs. (3) and (5) imply 
Using this result in cq. (7) fin&y yields 
L=F,or+F~*L+(6i(l~6)+F,)o~. (9) 
Fallowing arguments similar to those presented by Bruno (i968), suppose 
that a linearrelationship exists between the real martinal productivity of 
labor in cQ given sector and average output per Laborer in the economy, say 
F,=/V(Y/L). (10) 
Then, dividing eq_ (9) through by Y and denoting Fk =u yields, after some 
manipulation 
-+- [” ,I ! + 8 + FJ l (x/X) . (Xl Y). (11) 
The formulation in eq. (11) will be the basis of the empirical work reported 
in t;?e next section? Note that if marginal productivities are equalized across 
sectors (S =O) and if there are no inter-sectoral externalities (F, =O), then eq. 
(I 1) reduces to the famihar neo-classical formulation of the sources-c&growth 
mod& In the more g,eneral cm_=, the term [S/(1 + (i) + FJ is likely ta be non- 
zero for less d~~e~~~d countries. 
~~~~~t~er secti on will discuss a b-ecificatjon allowing a separate estimate of’ F,. 
C. Fee m, Ori exports and economic growth 63 
Under the present forrmlation the parameter cz should be interpreted irs 
the marginal productivity cbf capital in the non-export sector, rather than as 
nai productivity 8 capital in the economy as a whole. 
total increment to GDP brought about by a 
allocated to the export sector; TMPK, can be 
arginal productivity of investment in exports. 
Similarly denote by TNPLX the effect on GDP of a marginal increase ?n 
exprt sector labor. Then, c ne can shctw 
(TbfPL,- F,J’GL=(7’MPK,- F,)IG,==S/(l +a)+ I;,. (12) 
Eq. (12) clarifies the interpretation of the term on the extreme right-hand 
side (w&h in some of the empirical qork reported below will be estimate 
as a fixed parameter): it measures the difference between the marginal 
contribution to GDP of prduction factors in the two sectors, &c’ ative to the 
marginal contributions oft lese factors to export sector’s output. 
The interpretation of tte sources-of-growth equation [eq. (1 l)] is then 
str;;Fightforwarrl: the rate of growth of GDP is composed of the contribution 
of factor accumulation (i.c., growth of capital %nd labor) and the gains 
brought about by shifting factors from a low productivity sector (non- 
exports) to a high real produi=tivity sector (exports). 
Emmnct~ic formulation 
Eq. {I I)1 was used fcbr a c-ass-country regression relating the rate of growth 
of GDF’ (in constant price;) to thr share of investment in GDP, growth of 
population (proxy for labor growth) and to the growth of exports (in constant 
prices) multiplied by exports share in GDP. 
The study focuses on f. group of less deveLoped countries defined by 
Chenery (I 980) as semi&t ustrialized. The defini:tion involves both relative 
and absolute indicators (swh as the share of industrial output in GNP and 
the level of per capita i idustrial outp . SiEce the indicators do not 
necessarily overlap, Chene r~/ (1980) disti uishes between those countries 
which ire only ‘marginal”. y’ semi-industrialized, and those which qualify 
under a stricter definition. Accordingly, the study provides esti 
sample de.fined by the strict definition (19 countries) as well as a larger 
sample (3 1, observations) in:1 uding also marginal cases. 
An issues which needs to be addressed is the length of time to 
by any single country observation. Earlier s 
since annual data include substantial rand 
64 
annual .data are used. This study uses averages defined over the decade 19&b- 
fitj73,p 1 ” - : -i : ‘- : . 
There are ~>~host 4 ec;nometric I problems~ related to tiwwcount~y 
aggrQ4kte undy& .of grbwfh $q+,: poss~bihty zof sisfliul$aneous d4$wmin~thn 
of- both &perUcnt-.&ad ew#matory L v&iriBles~~ Wm a of these: oompkxtions 
are &mwxk by : ~Hagen’ can-d: 43a~ry4ysh.yn $I%?) ::and by Chenery et al. 
(3970), and will not be r4qMed here, +I& 4%8nsta’ti~, of parame&rs iacross 
observations merits, ho’pc. ever, some ftirtkr - elaboration. Any crsss-country 
study assur,mies implicitly that parameters in some generti way similar 
across coun$ries. In a production &nction eoctext, diffcrentxcountries are 
thus assumed to operate with identical production ,functions. In studies 
where the production funclion framework is complicated further by the 
existencx of non-optimal allocation, an additiijnal assumption is involved, 
namely,, thiat tiie degrkk of &s$ioc*at@n [&,ir&l%ated by &!-right-hand side 
of eq (12)] is similar &cross countries. # 
‘these j)robkx&s ri&d tr3 be borne in mind when parameters are 
interpreted. It is probably better to treat the est&atd coefficients as average 
values which provide a genera? order of magnitude *thin the sample but are 
not applicable to any specific country. An attempt :,lriill lx- made in a latter 
section to allow for a possibility of variation across observations in the 
externality effect F,. We set out now, however, to estimate eq. (1 I) in the 
form 
wherG the parameter y represents the differential pr~oductLities of factors, as 
explained earlier. It is expected that a;,the m+trginal productivity of capital in 
the non-export sector, will be positive but it’s magnitude should be less than 
the filpres in the range 0.24 35 which were reported by earlier studies4 
ba%ed on an aggregate m’acro pr&uction functioxl. The reason is that in such 
stt4dies the estimated parameter is some average (although not necessarily 
properly weighted) of gnarginatl ! l oductivities in the two sectors, and is thus 
!ikely to be higher than a spe- i W estimate of marginal productivity in the 
Xow productivity sector. The hypothesis that marginal productivities in the 
export sector are higher and that exports generate beneficial externalities 
suggests that the parameter ;I should be positi?re and significantly different 
from zero. The parameter ,/I!* related to labor growth, should aiso be 
cantly more than 2Ero if labor surp us we&s n!Dt the prevalent situation 
mple ~~~~~~~~~ ~~r~~~~ t’ic period covered. 
3. 
6. Z’ede+., On exports and economic growth 65 
ports the results af two specifications of the regression equation: 
rred to as tfre conventional neo_ctassicaI model, assumes that y =0, 
nts GDP growth as the result of capital and labor growth 
nd specificaaion follows eq. (13). Comparison of the two 
iights, therefore, the superior explanatory power of eq. (13). 
For both mmples, the acjusted Rf is almost doubled when the specification 
fur differences in I rginal productivit%s is used. T results lend 
pport to the h:po is that marginal factor prod ivities in the 
exPort sector are higher th;i.n in the non-export sector, as the coefficient of 
(&5’Q (X/Y) is positive anca significantly different from zero. The sign and 
magnitude of the co&cient related to investment are within the range 
expected. Specifically, when the conventional neo-classical fornlulation is 
used, the estimated parameter is within the range observed in earlier studies. 
When the formulation of eq. (13) is adopted, the parameter declines sharply. 
The parameter asso& ted Nith labor growth (which, as explained earlier, 
reflects the relation between marginal labor productivity and average output 
per laborer‘/ is significantly greater than zero.5 This may be taken as an 
Table 1 
Regression resul s for semi-industrialized LDCs, 1964 14-3 a 
- --- _-- - ~_- 
Extended sample Limited samp!e 
(including marginal cases) (excluding :.Darginal cases) 
- _- 
Variable Conventional net -classical Conventional neo-classical 
(parameter) model Eq. ( 15) model Eq. (1% 
Z/Y (a) 0.284 
(4.311) 
L/L (fl) 0.739 
( 1.990) 
(S/X) * w Y) iv) 
-2 R 0.370 
No. of 
obwvations 31 
--- - -----_ --_ - __-__ 
‘Numt)ers in parentheses are l-values 
0.178 
(3.542) 
0.747 
(2.862) 
0.422 
I( 5.454) 
0.002 
(0.180) 
0.689 
0.311 
(2.973) 
0.853 
( 1.652) 
-0.016 
(0.611) 
0.33 P 
0.196 
(2.432) 
4x737 
( t .9761 
0.390 
(3.985) 
0.0 
10 
__ -__- __ - 
‘With a !5”/:, one-tailed test. 
indication that surplus labor was not the general case fm sample countries in 
the period under investigation. 
Wide F& and y we;n: estim&d, Fs is t)iot knyvn (ia a iatm section of this 
chapter s&me direct estimates, sf Fi wU1 be:oI$&ti&i).For a.,pIausibe range 
of & iarlues, one ~a& however* c&uIa~e ‘@e :$otiespunding valves of the 
so&i marginal product of investmen\ ih exparts+- using eq. (14). These are 
presented in tabfe 2. 
0.10 0.289 0.303 
0.20 9.275 0.290 
0.30 0.264 0.280 
0.80 0.255 0.272 
0.50 0.248 0.265 
1.00 0.226 0.243 
-- - 
It &ouM be emphGzed that the estimates reported in table 2 are stricly 
con&tent for only one particular (and unknown) value of FXa This follows 
from the fact that y incorporates &, and thus varying the latter while 
holding y fixed is not strictly le,$timate. However, it is evident from these 
calculations that the marginal vallre to the economy from a unit investment 
in exports expansion is sebstantial$ higher than that of investment in non- 
exports. While these numbers co/ .~~pond to sample averages, they provide 
strong support to the vie \ri that a’ ,z success story of export-led economies 
such as Korea is due in large pdrt to the enormous shift of resources into the 
higher productivity export sect~t. 
expor’s affix the production of non-exports with consrant elasticity, i.e., 
N = &K,, LB, X) = x@ * $(K,, LJ, il5) 
where 8 is a 1 arametw. One can show 
~~/~~~ sz F, = 0 l (N/X ). 
Eq. (11) caq now be rewritten 
(16) 
P 1 t I”’ x x -= 
y 57 l --+8_z+ 1 ,k6 
_.L+&__L._._ 
x x Y' 
But 
[‘sing this result, eq. (17) is rearranged, obtaining 
Note that if it is assumed a/(1-,- 6) = 0, the model reduces to 
which is essentially the equ 
Balassa (1978) and Tyler ( 198 l)? 
adopted by Michalopeulos and Jay (1973), 
Results of regressions adopting the specification of eq. (18) are reported in 
table 3. 
The modified formulation increases the explanatory power of the model 
considerably. The results indicate that the nter-sector externality parameter 
(e) is statistically significant in both samples. The magnitude of the estimated 
parameter is quite substantial. If exports are increased by 10% without 
withdrawing resources from the non-export sector, rhe latter grows by 
approximately 1.3”/,. The other component of productivity di 
rameter 6) can be calculated ‘verr the estimate of 0 and t 
with (JQX) - (X/?I). T 
1 productivity diaerential 
addition to the differential ue to externafities. 
Ta’ble 3 
Regression results for semi-industrialized LDCh with specific irate:-se&ral externalities, 1964- 
1973.” 
-,-c 
wuiabie 
(parameter) 
(a) 0.124 
(3-O@) 
L/L 1(19) 0.696 
(3.399) 
(X~‘x)~(X/Y)(S/(l +J)--8) 0.305 
(4.571) 
g/x (8) 0.131 
(4239) 
Constant 0.006 
W.596) 
-_ 
R2 0809 
No. of 
&ervations 31 
‘Figures in parentheses are t-values. 
0.507 
(L921) 
o.goz 
(3S81) 
0,124 
(2.989) 
0.005 
(3.331) 
,, 
0.773 
19 
The coefficients of investment share and labor growth are ,>within the - ne 
order of magnitude of the estimates obtained in table 2, and are statist@* ally 
sign&ant.’ 
Using the results of table 3, the social marginal product of investment in 
exports (TAM%,) can be calculate& Recall from eq, (16) F, = 0 l ((I- X)/X), 
ywhcre x is the share of exports in GDP. It follows then, 
Yfx) 
I-x =- lb0 l - x ’ 
and eq. (14) can be written 
(19) 
Using the parameters repo~,-_,J in table 3, eq. (19) generates the social 
marginal product of investment in exports for economies with different values 
of X. These are presented in table 4. 
The calculations in table 4 are free of the inconsistencies associated wit4 
the numbers presented in table 2. Again, it is demonstrated i that, at the 
margin, investment in the export sector has a substantially higher s~%l 
Table 4 
Social margina product of investment in exports semi- 
industrialized LDCs, l%4-1973. 
Share z$ exports 
in GDP (xj 
0.1 
cl.15 
0.20 
0.25 
0.30 
0.35 
0.40 
0.45 
0.50 
-- 
TMPK, 
Extended sampk 
0.439 
0.383 
0.335 
0.306 
0.287 
0.273 
0.263 
0.255 
0.248 
Limited sample 
0.512 
0.412 
0.362 
0.332 
0.312 
0.298 
0.287 
0.279 
0.272 
marginal productivity thcrn investment in non-exports. The differentid is 
smaller in economics where a large share of resources is already allocated to 
the export sector, as in such economies the inter-sector externality effect is 
smaller. 
3.3. Sources of growth 
Using the results of table 3 anti the mean values of sample variables, th:: 
average rate of GDP growth in the period f 9644973 can br: decomposed so 
as to identify the contributions of aggregate investment, latter force growth 
and r-esource shifts into exports. This is done in table 5 for both samples. The 
resAts indicate that the gain in productivity due to closer-to-optimal 
Table 5 
Sources of growth of semi-industrialized LDCs, 1964-1973. 
--_- -w -- -- 
(2) (3) 
r, 
(1) Parameter Contribution to gr~th 
Mean in ,ample (from table 3) 10) x(2)1 x 100 
.-- --- 
Extended Limited Extended Limited Extended Limitc:d 
Variable sample 
_--- 
l/Y 0.205 
i/L 0.02 3 
(x’/X) * (Xl Y) 0.022 
X/X 0.084 
constant 
GDP growth 
t F/Y) 0.065 
sample 
0.222 
0.020 
0.03 
0.105 
sample sample sample samp t 
________ ____~__ 
0.124 0.139 2.54 3.09 
0.696 0.587 1.60 1.17 
0.305 0.302 fO.h7 0.9 1 
0.131 0.124 I.10 1.30 
0.006 O.005 0.60 0.53 
‘I.070 4.5 7.8 
--_-- --___-_- -___ ______-______ 
D 
allocation associated with export expansion contributed more than 2.2 
--1*4-?a CT- )AilS#W.B,ugv points to, the. growth -of -the s!,k&ly -sen&industriali countries, 
and close to 1.8 pera&ge *’ p$ii-ts if i the’ ’ less strict definition of 
industrialiition~ is used, The tinfr%utiog of exports car. be decomposed into 
two compormen~s: (i) Tne gain due to &&&Sal extqrnalig.ies affecting the 
non-export sector- &&ich ‘equals L’ 19 l (I -x) . @%/.X)J: (ii) the: gain due other 
elemeirats undcHying higher factor proc@tivity in the export sector 
((W(I +4)*WX)*(X/y)). 
The calcuMions for the extended sample reveal that 0.81 of one percentage 
point is due to inter-sectoral externalities, and 0.96 is due to other efkcts. 
The Correspondiag figures for the limited sample are 0.93 and 1.27. Thus, 
slightly less than half of the gain.in growth due to higher factor productivity 
in exports is {due to inter-sectoral externalities, 
It is of interest to apply the specification of eq. (18) to a sample of 
developed =onomies so as to test for the existence of marginal factor 
productivify differentials and qxternaligy effects. The estimatc:s are reported in 
table 6 for both t& neo-classical formulation and eq. (i8); The results suwest 
that there is a substantial externa.lity elect, but the extent of marginal factor 
productivity differentials can not be established, since the coefficient of 
(x/x)*(X/Y) is not statistically significant. Using the point estimate as a 
basis for calcula’tion yields S/( I+ 6) = 0.25, which implies S = 0.33. 
The explanation of the significant difference in the estimated magnitude of 
the parameter 0 among developed and developing countrie:; requires more 
detailed analysis, which is beyond the scope of this paper. 
Table 6 
Regression results for d xeloped economies, 1964-1973. 
Variable Nr:o-classical Model with externality 
(parameter) ?mnulation effects 
-- 
l/Y ta) - 0.350 0.1408 
(? 26) (2.865) 
i,;L (8) 4 .( ‘05 0.6595 
I LJ12) (1,483) 
cS,‘x)~cx/Y)(a/(1+6)-8) - Q,2400 
(1.308) 
X/X (0) CA4938 
(5.477) 
Constant - 0.0242 - 0.0301 
(1.234) (2.608) 
L 0.444 0.815 
17 J7 
-_- ----_ 
luding tezarks 
This paper provides 
economies which adopt 
71 
evidence supporting the view that the success of 
export-oriented policies is due, at least partially, to 
the fac,t that such policies bring the ezonomy closer to an oplimal ak~ion 
of resources. The estimates show that there are, on averzkge, substantial 
differences in marginal factor productivities between the er.port and non- 
export sectors. These differences derive in part from the failul e of 
entrepreneurs to equate marginal factor productivities and in part dJe to 
externalities. The latter are generated because the export sector confers 
positive effects on the roductivity in the othersector? bul these ars not 
refiected in market prices. 
The results are such that social marginal productivities are higher in the 
export -sector, and economies which shift resources into exports wil; gain 
more than inward-oriented economies. The empirical findings suggest that 
even when entrepreneurs optimize resource allocation given the prices they 
face, there are substantial gains to be made due to the extemz.lity effects. 
The analytical framework developed in this study can be u%ized in studies 
using more detailed data such that the extent of productivity differential in 
specific groups of countries (with dSfcrz:lt policy orientation) can be assessed. 
Similarly, the relation between inter-sector externalities and export 
composition can be clarified further using the same analytical framework. 
Appendix: Sources of data and definitions 
.A. I. Cciculation of variables 
All data were obtained from World Tables 1980. Variable: were calculated 
from time series for the period 19661973 in constant prices. Average rates of 
growth were obtained by regressing In 2,. -a + be t where Z, is the economic 
variable under consideration and t is time. The rate of growth. say r, is then 
calculated as r =eb - 1. Average ratios (investment/GDP, export :GDP) were 
calculated as simple averages for the decade. 
Table A.1 presents means and standard deviations of the variables wed in 
the study. 
.4 2. Cmpnsition of samples 
Following Chmery (1980), the st* ict definition of semi-irndustk-ialized LDQ 
e excluded from 
ADEILSON
Highlight
ADEILSON
Highlight
72 G. Fedttr, Qn exports md ecopromic growth 
Population growth 
Exparts growth 
Export share 
(Export growth) x 
e ;porP share 
0.023 0.020 
(Q.809) (Q.009) 
0.084 Oil05 
(0.070) (0.080) 
0.235 0.266 
(0.225) (0.282) 
0.022 0.030 
(0.032) (0.038) 
t!ie forrr,er is defined by the World Bank as a developed country and the 
latter is a major oil exporter. Countries which are considered as marginally 
semi-ir~dustrialized were added to the limited sample creating the ‘extenfied 
sample’. These countries are: Dominican Republic, Ecuador, Egypt, 
Guatemala, India, Ivory Coast, Kenya, Morocco, Peru, Philippines, Syria, 
Thailand and Tunisia. In addition, Iran, Iraq and Algeria are defined as 
nzarginal GHB but were excluded from the sample, being major oil exporters. 
Tht: samplk? of developed (countries consists of Australia, Austria, Belgium, 
Cana&% Dennlark, Finland, Fr mce, Germany, Ireland, Italy, Japan, The 
Nethenands, Norway, Sweden, SCtzerland, U.K., and U.S.A. 
Bdassa, Bela, 1977, Policy r&x ZI in u(:velt)ping countries (Pergamon Press, Near k ork). 
Balassa., Bela, 1378, Exports and economc growth: Further evidence, Journal Df Development 
Economics 5, no. 2, .:ux, 181-189. 
hagwzti, Jagdish N. a(lcA T.N. Sriniv;isan, 1978, Trade policy and developm cnt, ira: Rudiger 
Dombusch and Jacob A. Frenkei, eds International economic policy: Theory and evidence. 
ch. 1 (Johns I-Iopkins Uti~t~~ity Pim, Baltimore, MD) l-38. 
rxsno, Michael, 1968, Estimation 3f factor contribu ion to growth under structural 
nte~nationa~ Economic Review, Feb. 
~9g~, The ~r~~j~~~dustria~ count*.ies, Draf,, July (Worhi Eaak, Washington. 
azeiI F~ki~lgton and Christopher Sims, 1970. A uniform amlysis of 
s, Eix mnic DeveTopmePt report no. 148, July (Ktrvsrd University, 
H~gen, E. Everett and 0li Hawrylyshyn, 1969. Analysis of worl,;l mcome and growth. I ~~55 
15365, Economic Development and Cultural Change 18, no. 1, 
Weller, Peter S, aid Richard Porter, 1978, Exports and growtf 
Journal of ikvelopment nomics 5, no. 2,5me. 191-19?. 
Humphrles, Jane 1978, Causes of growth, Economic Development and Cultural Change 24. no. 
2, Jan., 339-3’53. 
ibzsing, Donald B., 1967, Outward-looking policies and economic developmer.t, Economtc 
Journal 77, no. ,306, June, 303-320. 
Keesing, Dunafd B., 1979, Trade policy for developing countries, World Bank Staff working 
paper no. 353, Aug. 
k-avis, Irving B., 1070, Trade as a handmaiden of growth: Similarities between the nineteenth 
and twentieth centuries, Economic Journal $0, no. 320, Dec., 8% “72. 
Krueger, Anne 0.. 1980, Trade policy as an input to development, American Economic Review 
70, no. 2, May, 288-292. 
Michaely, Michael, 1977, Exports and growth: An empirical investigation, Journal o 
Development Economk 4, no. 1, March, 49-53. 
Michalo,mulos, Constantine and Keith Jay, 1973. Growth of exports and inm be in th : 
developing world: A neoclassical view, AID. dkussion paper no. 28 (Washington, DC). 
Pesmazoglu, John, 1972, Growth, investment and saving ratios: Some iong and me&urn term 
associations by groups of countries, Bulletin of Oxford University Institute of EcontJmics and 
Statistics 34, no. 4, Nov., 309-328. 
Robinson, Sherman, 1971, Sources of growth in less developed countries. Quarterly lournal cd 
Economics 85. no. 3, Aug., 391-40@. 
Tyler. William, 1981, Growth and export expansion in developing c