P_S
Journal of Phillpplne Developmen t
Number Thirty*Two, Volume XVIII, No. 1,19g 1
P_
S
ECONOMIC GROWTH AND DEFENSE SPENDING:
EVIDENCE ON CAUSALITY FOR SELECTED
ASIAN COUNTRIES*
P. C.Frederiksen* *
1. INTRODUCTION
In the lastdecadeincreasingattentionhasbeenfocusedonthe role
of militaryexpendituresindevelopingcountries.Onereasonfor this isthat
worldmilitaryexpendituresin 1987 exceededonetrillionUSdollarsfor the
veryfirsttime;developingcountriesaccountedforapproximately17percent
ofthistotal.Despitethesteadygrowthindefensespending,scholars,forthe
most part, know little aboutthe effects of militaryspendingon the local
economy. As Saadat Deger(1986: 3) recentlynoted: "The numbersare
mind-boggling;...Butwhat isclearisthatitisessentialto study,analytically,
theeconomicdimensionsof militaryexpenditureinlessdevelopedcountries
(LDCs) and carefullyevaluate the costs and benefits involved.... the
economicsof militarizationarecrucial."
Anotherreasonforthe increasedawarenessof militaryexpenditures
isthat developingcountrieshavehadto recentlyscrutinizethesizeofoverall
budgets, especiallythe defensecomponent.In manycasesthe resource
base has declinedbut the needsfrom othersectorsof the economy have
concomitantlyincreased.In manyinstances,defenseexpenditureshave
replacedhighprioritydevelopmentprojectswith littleknowledgeonthepart
of governmentsasto the impactof sucha decision.
Much ofthecurrentresearchin thefieldof defenseeconomicscanbe
attributedto the early work of EmileBenoit(1978) who concludedthat,
contraryto popularopinion,developingcountrieswhichhadhigherdefense
burdens (defense spendingas a percentageof Gross National Product
(GNP)) usually had higher rates of economic growth. This result was
diametricallyoppositeto the usually-heldbeliefthat increasesin defense
spendingmeantlowergrowth;scarceresourceswere siphonedaway from
moreproductiyeuseselsewhereinthe economy.
*An eadier version of this paper was presented in 1989 at the Institute for Strategic and
InternaldonalStudies, Kuala Lumpur. I wish to thank two anonymous referees formaking valuable
sugge_tJor_.
"° Professor of Economi_s, Naval Postgraduate School, Monterey, California, USA.
132
JOURNAL
OF PHILIPPINE DEVELOPMENT
In the last decade, other aspects of defense spending in developing
countries have received closer attention. Some of these areas are:
-
howmilitaryspendingaffectsratesofsavingsandinvestment,
and
consequently prospects for future growth;
the economic and noneconomic determinants of defense spending;
the determinants of arms industries;
the production of major weapons systems by only certain developing countries;
the discernibleeffects (both economic and noneconomic) of civilian
versus military regimes; and
human capital formation and development in the military. 1
In a recent paper, Looney and Frederiksen (1990) examined the
determinants of defense spending in six Asian countries: the Philippines,
Indonesia, South Korea, Malaysia, Thailand, and Singapore. Their research
suggested that economic variables and resource availability were the main
determinants of military expenditures inthese six countries. The purpose of
this paper isto test, for the same set of countries, the widely-held hypothesis
that the direction of causality is from military expenditures to economic
growth, i.e., that defense spending can in fact "cause" economic growth.
The alternative hypotheses are that (a) economic growth precedes defense
spending, thereby allowing countries to increase defense outlays; (b) no
relationship exists; or (c) a feedback loop exists whereby defense leads to
growth which, in turn, leads to more defense, etc. The working hypothesis
for this paper is that, aprior/, one cannot specify the direction of causality
for any individual country. In addition to testing the usually assumed
hypothesis of defense to growth, this paper goes on to (a) extend the
preliminary results obtained by Frederiksenand LaCivita (1987) on causality
in the Philippines which originally appeared in this Journal; (b) extend the
results obtained in LaCivita and Frederiksen (1991 ) from 1982 to 1988 for
the Philippines and Thailand; and (c) include comparative research results
for Indonesia, 2 Malaysia, Singapore, and South Korea.
THE CAUSALITY
ISSUE
Most studies to date have assumed that defense spending is an
exogenous variable and, thus, "causes" economic growth. In other words,
the models which have been tested have specified economic growth as the
1. For an excellent review see Deger (1986) and Lindgren (1988).
2. The results for Indonesia converting the period 1864-85 ere reported in Frederiksen
(1989). The results for Indonesia in this paper cover the period 1961-88.
FREDERIKSEN:GROWTH AND DEFENSE SPENDING
133
dependent variable. Governments undertake military expenditure programs; and, presumablyat some (unspecified)time in the future, the
economywould eitherbenefitorsufferasa resultofthis spending.Most of
the studiesinthe1980s focusedonwhethertheeffectwas positive,neutral,
or negative(seeDeger1986). Thewidely-heldbeliefregardingtheexogeneity
of defense spending originated with Benoit. In his seminal paper, he
recognizedthat the directionof causalitycouldbe eitherfrom defenseto
growth orfrom growthto defense.As Benoitnoted:
A questionarose,however,aboutthedirection
ofthisinteraction.
Mightnot
the correlation
be explainedby the influenceof growthrateson defense
expenditures
ratherthanviceversa?Countries
withrapidgrowthmightfeel
better able to indulgethemselvesin the luxuryof elaboratedefense
programs
.... Theseplausible
hypotheses
didnotappeartobesupported
by
theevidence
.... Thusgrowthappearedto exertonlya weakinfluenceon
defenseburdens.(Benoit1978:275-76)
The assumedexogeneityof defensehasrecentlybeenchallengedby
Joerding (1986) who examineddata for 57 countriesandconcludedthat
defenseexpenditureswere notstronglyexogenous.Whetherornotdefense
spendingpromoteseconomicgrowth is importantfor those developing
countriessearchingforwaysto improveeconomicperformance-orat least
to spendsoasnotto retardeconomicgrowth.Joerding'sworkis undoubtedlyan importantcontributionto the literature.Butas was pointedout in
FrederiksenandLaCivita(1987)andLaCivitaand Frederiksen(1991 ), there
are two major criticismsof Joerding's work. First, Joerding lumps all
countriesintoonesample.Thissuggeststhat ifa causalrelationshipexists
it iscommonto allcountries.Itisquitelikelythat in somecountriesdefense
causes growth, in others growth causes defense, and yet in others a
feedbackrelationshiporevennorelationshipexists.The secondpointisthat
Joerdingassumeda commonand arbitrary four-year lag structureonthe
defenseand growth variablesin hispooledset. It isreasonableto assume
thatthelagstructuremightdifferfromonecountryto anotherdependingon
the structureof the economy,thetype of defensespending,and the like.
In Frederiksenand LaCivita'sstudy (1987) of the Philippinesduring
1956 and1982they usedJoerding'sprocedure(explainedmorefullybelow)
totestforcausality.Nostatisticallysignificantrelationship
was foundto exist
betweendefenseandgrowthwhenthelaggedvaluesforthevariableswere
arbitrarilychosento befour years.Whenthe modelwas re-estimatedwith
anarbitrarylagvalueoftwo, itwasfoundthatgrowthcausesdefense.These
initialresultssuggestthat, sincebothof thelagstructurestested (two and
four years) were chosenarbitrarily, one cannot say which isthe correct
specification.
134
DATA
JOURNAL
OF PHILIPPINE DEVELOPMENT
SOURCES
The primary data sourceSon Gross Domestic Product (GDP) and on
the rate of growth of realG DP (except for Malaysia between 1961 and 1970)
was the International Monetary Fund's (IMF) internationalFinancialStatistics Yearbook, The Yearbook: 1984 was used for data up to 1959, the
Yearbook: 1990fordata between 1960 and 1980, and the Yearbook: 1991
for the period 1981-88. Defense expenditures were drawn from the United
Nation's Statistical Yearbook (annual issues through to 1983/84). The rate
of growth of real GDP was not reported by the IMF for Malaysia between
1961 and 1970. A series was constructed using the nominal GDP adjusted
for inflation/deflation by the Consumer Price Index (CPI). The defense
burden through to 1983 (for all countries except Singapore and Indonesia)
was calculated as the percentage of GDP allocated to defense expenditures. The defense burden data for the remaining years (1984-88) were
reported by the United States Arms Control and Disarmament Agency
(ACDA) in World Military Expenditures and Arms Transfers (annual issues).
A comparison of several overlapping years (Appendix A) indicated similar
ratios for the military burden whatever the source used .4Defense expenditures for Malaysia between 1969 and 1985 were taken from the Institute of
Southeast Asian Studies' Defense Spending in Southeast Asia, (1987). The
defense burden for Singapore was derived exclusively from ACDA data.
The defense burden for Indonesia for the period 1961-66 was reported by
the Stockholm International Peace Research Institute's (SIPRI), S/PRI
Yearbooks on Wor/d Armaments and Disarmament. The remaining years
were drawn from USACDA.
The following sections describe, respectively, the methodology by
which the optimal lag length and correct direction of causality were calculated andthe empirical results for the Philippines(1956-88), Thailand (195688), South Korea (1955-88), Malaysia (1961-88), Singapore (1967-88), and
Indonesia (1961-88).
METHODOLOGY
Hsiao (1981 ) has developed a systematic method for choosing lag
lengths to avoid the problems associated with arbitrary lag lengths. His
method combines Granger causality and Akaike's final prediction error
(FPE). Initially a series of regression equations is estimated on the depend-
3. The data used in the study appears as Appendix A, Tables A 1-A6.
4. Sinoe the tests aremore robust with longerperiods, we feel more than justified in using
ACDA data on the military burden for the later years especially given the similarity to the ratio
computed (using UN and IMF data) for aadier years. A data sheet for each country, together with
a copy of the computer output, is appended to this paper.
FREDERIKSEN: GROWTH
AND DEFENSE SPENDING
135
ent variable. In the first regressionequation, the dependentvariable is
laggedoneyear, andin succeedingregressions,anadditionallagisadded.
That is, M regressionequationsareestimated in theform:
m
Gt = (x,+_, _t-, Gt-1+ st
i=-1
where Giseconomicgrowthand whereM, the maximumlaglength,takes
on thevaluefrom 1 to M.s
Foreachregression,thefinalpredictionerror(FPE)iscomputedin the
followingmanner:
T+m+
FPE(m) =
T-m-1
1
ESS(m)/T,
where Tis the samplesize, andFPE(m)andESS(m) arethe finalprediction
errorandthesumofsquares,respectively.Theoptimallaglength,m*, isthe
laglengthwhichproducesthesmallestfinalpredictionerror,i.e., themost
accurateforecast.Oncem* hasbeencalculated,anothersetof regressions
isestimatedwithlaggedvaluesofD, thedefensevariable,addedsequentially
inthesame mannerwhich was usedto determinem ° . Thus, six additional
regressionsareestimatedin theform:
rrf
n
Gt = oc+___PI-1Gt-1 + _._'Yt-lDt-I+
i=1
F_t,
i=1
with n takingon the valuesfrom one to six. We then compute the FPEfor
each of these regressionequationsas:
T+m* +n+ l
FPE(m* ,n ) -
T-m*
-n - 1
_ESS(m*, n)/ 7",
and we choosetheoptimallag lengthfor D, n *, asthe regressionequation
withthelowestpredictionerror.
In thesame manner,anothersetof equationsisestimated, with D as
the dependent variable,and laggedvaluesof D are includedto find m*.
Then, laggedvaluesof G areincludedto findn*.
5. Although the ohoioe of Mis arbitranh it should be as large as possible, consiatent with
the sample size and the u nderlying process. Becau so of the relative shortness of time series data
for mostdevsioping oountdes, M was limited to six years.
136
JOURNAL
OF PHILIPPINE
OEVELOPMENT
The final procedure is to test for causality, and this consists of three
steps:
STEP 1: The final prediction error for the model G = f(GL) iscompared
to the FPEfor the model G = f(GL, DL). If the FPEdecreases (i.e., the model's
predic'dvepower increases as we add lagged values of D), we conclude that
defense Granger causes growth. If, on the other hand, the FPEincreases
then we conclude that defense does not Granger cause growth.
STEP 2: The FPEfor the model D = f(DL) is compared to the FPEfor
the model D = f(DL, GL). If the FPE declines, we conclude that growth
Granger causes defense. If the FPE increases, we conclude that growth
does not Granger cause defense.
STEP 3: We compare the FPEs under Step 1 and Step 2. If the FPE
increased in both cases, we conclude no relationship between D and G. If
the FPEdeclines in both cases, we conclude a feedback relationship exists.
If the FPEdeclined under Step 1 but increased under Step 2, we find that
defense Granger causes growth. If the FPE increased under Step 1 but
declined under Step 2, we find that growth Granger causes defense.
EMPIRICAL
RESULTS
Lag Lengths
The optimal lag lengths using I-Isiao's method on each country appear
as Table 1 .eThe paired numbers in the first column indicate the optimum lag
lengths, m* and n*, respectively, when economic growth is the dependent
variable and lagged values of G and lagged values of the defense variable
are the independent variables. The paired numbers in the second column
indicate optimum lag lengths, m ° and n °, when defense is the dependent
variable.
As hypothesized, the estimated optimal lag lengths differ slightly
among the countries. While we are unable at this aggregate data level to
sPecify reasons for individual lag length differences among the countries, it
is comforting to see that most of the lag lengths are either one or two years.
These results support the findings of Looney and Frederiksen (! 990) who
looked at the determinants of defense spending in these countries.
Past growth of GDP has an immediate impact (one year) on current
GDP in four of the countries. The one anomaly is a six-year lag rate for
Malaysia, However, this result must be looked at with caution since Hsiao's
model, while relatively consistent, tends to overestimate lag lengths. The
6. The final predlotion orrors havo not been reported in the paper. As noted pre_Aously,
the results oan be obtained directly from the author.
FREDERIKSEN:
GROWTHAND DEFENSESPENDING
137
Table I
OPTIMAL LAG LENGTHS (Years)
• It
Country
G = f (GL , DL)
D : f (D, , G_
Philippines
(1, 1)
(1, 1)
Thailand
(1, 1)
(1, 2)
South Korea
(1, 5)
(1, 3)
Malaysia
(6, 1)
(2, 3)
Singapore
(2, 2)
(2, 1)
Indonesia
(1, 1)
(1, 1)
result for Malaysia, while not spurious, is probably less than the computed
six-year computed lag. In other words, earlier rates of growth of the GDP
have a more recent effect on current GDP growth than indicated by the
estimated results. In addition, the six-year lag isan average over a twentyeight year period and is most likely considerably shorter if examined using
only recent data.
Importantly, the data indicate an almost immediate effect of the
defense burden on growth: a one-year lag for all countries except Singapore
which experiences a two-year lag (see Column 1 ) and South Korea (fiveyear lag). In addition, the impact of past defense spending on current
defense spending is immediate (one year) in four of the countries, and two
years for Thailand and Singapore. This result also supports earlier research
findings on the determinants of defense spending. It appears as if the
"jumping off point" for the current military budget iseither the military burden
or the growth rate in G DP of the preceding year. Past GDP growth rates have
a one-year lagged impact on defense for the Philippines, Singapore and
Indonesia. Past GDP from two years ago impacts the current defense
budget in Thailand, and the impact is three years for South Korea and
Malaysia.
DEFENSE/GROWTH
RELATIONSHIPS
The final step in the procedure isto compare the final prediction error
for the regression equations to determine causality, if any. The relationships
appear in Table 2.
For a half of the sample, the data suggest a clear direction of causality:
from economic growth to defense for Malaysia and from defense to growth
for Singapore and Indonesia. For the remaining three countries there is
either no relationship between growth and defense (Philippines and South
138
JOURNAL
OFPHILIPPINE
DEVELOPMENT
Table2
DEFENSEAND GROWTHRELATIONSHIPS
Country
Relationship
Philippines
Thailand
No relationship
Feedback
SouthKorea
Malaysia
Singapore
Indonesia
No relationship
Growthto Defense
DefenseGrangercausesGrowth
DefenseGrangercausesGrowth
Korea) or a feedback loop (Thailand). As noted above, the aggregative
natureofthedataprohibitsusfromidentifyingthespecificdefenseexpenditure whichcausestheeconomicgrowthoreventhetimingof theeconomic
growthwhichinducesthefollow-ondefensespending.Thisisprimarilydue
to the examinationof data for a thirty-yearperiod.Presumably,for SingaporeandIndonesia,theexistenceofa largearmsindustry(especiallyinthe
case of Singapore)or of growth-inducingtypes of military spending(on
infrastructure,education,housing,andthelike)hashadpositiveeffectson
theeconomy.These aretheeffectsthat Benoitsuggestedmighttake place
in developingeconomies.ForMalaysiathefinal predictionerrorsuggests
that economicgrowth is an important determinantof defensespending,
whichisoppositetothe patternsuggestedby Benoit.ForThailand,thefinal
predictionerror declinedwhen G and D were the respectivedependent
variables.Thissuggestsa "feedback" relationshipwherebygrowthhasled
to moredefensewhich, inturn, hascreatedmoregrowth, and soon. This
was the predominantcasein the largerstudy conductedby LaCivitaand
Frederiksen(199i ). Fortwo of the countries,the Philippines'and South
Korea,theresultsindicatenorelationshipbetweendefenseandgrowth. For
Koreathis issomewhatpuzzlinggiventheoftenperceivedimportanceofthe
defensesectorineconomicdevelopment.However,thisresultmightbedue
to the inclusionof relativelyrecentdata (throughto 1988) in the model;
Korea's growth in the last decade or so may no longerdepend on the
defense sector (as much as it did earlier), and other factors (such as
technological advances and economic diversity) have been the main
enginesofgrowth. Forthe Philippines,asinthecaseof Korea,theinclusion
of the defenseburdenasanindependentvariable(with Gasthe dependent
variable)ortheinclusionof G asan independentvariablewhen D was the
dependentvariableled to increasesin the final predictionerror; defense
expendituresinthePhilippinesapparentlyhavenoGrangercausalityeffect
FREDERIKSEN:GROWTH AND DEFENSE SPENDING
139
on economic growth, and growth doesnot"cause" defense. Presumably
either pattern couldhave happened(or is happening)if we were to take
shorterperiodswithin thethirty-yearperiod.Unfortunatelythe shorter the
time periodexamined,the weaker isthestatisticaltest.
Froma policypoint of view, plannersin South Koreaand the Philippinesshouldthereforenot rely ondefenseexpendituresasa majorpolicy
toolforeconomicgrowth intheir respectiveeconomies.The implicationof
theseresultsisthat anyeffort to imposearbitrarylaglengthsalongthelines
of Joerdingon modelswhichtry to uncovercausalitywilllikelyhidethetrue
relationship.The appropriaterelationshipcanonlybedetectedby a proper
specificationofthe model,andthe appropriaterelationshipWillmorethan
likelydifferfrom one countryto another-an assumptionnot felt likelyby
earlierauthorssuchas Benoit.
SUMMARY AND CONCLUSIONS
This paperhasexaminedthe defense/growthcausality issuefor six
Asian countries. The purposewas (a) to extend an earlierpaper in this
Jouma/dealing with lag lengths in the Philippines;(b) to update prior
causalitystudiesonthe PhilippinesandThailandthrough 1988; and (c)to
includecomparativestudies on Indonesia,South Korea, Singapore,and
Malaysia.
Most past studieshaveassumedthat defensespendingprecedesor
"causes" economicgrowth-i.e., thatcorrelationimpliescausation.Joerding
suggestedthat defensespendingisnotexogenous.The two criticismsofthe
Joerdingpaperwere that (a)he useda pooledsample,and (b) he chosean
arbitrarylagstructure.
Thissix-countrystudy determinedona country-by-countrybasisthe
optimallagstructurefor thedefenseandgrowth variables,ontheone hand,
and the appropriatedirectionof causalityfor each country, on the other
hand.
The results indicate that the lag structure differs from country to
countryashypothesized.However,whenbroadlyexamined,thelaglengths
areprimarilyone ortwo years. The similarityisfurther strengthenedwhen
oneconsidersthat the modelusedin this papertends to overestimatethe
lagperiod.Pastgrowth andpastdefensehavean almostimmediateimpact
oncurrentgrowth, andsimilarly,pastdefenseandpastgrowth havea very
quickimpacton currentdefense.
The results alsoconfirmedthat the causalrelationshipdiffers from
country to country. ForSingaporeand Indonesiathe resultsindicatethat
defense Granger causes defense-the presumed directionaccording to
Benoit.Onlyin Malaysiadideconomicgrowth appearto be a determinant
of defensespending.In the caseofThailandourresultssuggesta feedback
140
JOURNALOF PHILIPPINEDEVELOPMENT
relationship; for the remaining two countries-Philippines and South
Korea-no discernible relationship between defense and growth was uncovered.
The Irnpllcatlon
forfurther
researchisthatneither
defenseor growth
can be consideredas exogenously determined.Furthermore,
we do not
expectone model tofit
allcountries
oreven a group ofcountries
Inthesame
region.
The laglengthsarelikely
to differ
from countryto country,as does
thecausalrelationship
between thedefenseburdenand economic growth.
Given thecomplexityoftheinteraction
ofthedefenseand growthvariables,
a fruitful
areaforfuturereseamh mightbe to uncoverexactlyhow defense
affectsgrowth on a count_/-by-country
and year-by-year
basis-a task
made difficult
due to the aggregativenatureof the reporteddata.Given
further
breakdowns as to thetypesof military
spending{on,forexample,
capital
equipmentor R&D), more completeinsights
mightbe uncoveredas
to theroleofmilitary
expenditures
on theeconomy.
APPENDIX A
Table A1
DATA, REPUBLIC OF THE PHILIPPINES
Year
Gross
Domestic
Product=
Defense
expenditureb
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
10
11
12
13
14
15
17
20
21
23
26
29
32
35
42
50
56
72
1O0
115
166
157
181
181
190
197
206
209
230
225
249
290
345
326
458
543
602
855
1941
3982
Defense
USACDA
Growth rate
burden
defense
of real GDP
(pement) burden (percent) (percent)
1.6
1.4
1.5
1.4
1.4
1.3
1.2
1.0
1.1
1.0
1.0
1.0
1.1
0.9
1.1
1.1
1.1
1.2
1.9
3.5
6.9
5,3
3.5
6.7
1.5
5.6
4.8
7.0
3.5
5.2
4.4
6.1
5.6
4.8
4.6
4.9
4.8
9.2
5.0
6.4
FREDERIKSEN:
GROWTHANDDEFENSESPENDING
141
Table A1 (continued)
Year
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
t988
Gross
Domestic
Producta
Defense
expenditureb
135
,154
178
218
265
305
341
384
4118
4325
3552
4995
5115
5526
5552
6106
Defense
USACDA
Growth rate
burden
defense
of real GDP
(percent) burden(percent) (percent)
3,1
2.8
2.0
2.3
1.9
1.8
1.6
1.6
2.0
2.3
1.9
1.8
1.7
1.6
1.2
1.3
1.9
1.8
1.7
8.0
6.1
5.5
6.3
5.2
3.2
3.6
1.9
-7.6
-7.4
3.4
4.8
6.3
a. Billions
ofpesos.
b,Millions
ofpesos,
Sources:GDPand GrowthRateof RealGDP:IMF,InternationalFinancialStatistics
Yearbook:1984(1956-1959),
Yeadoook:
1990(1960-1980),
Yea_ook:1991(1981-1988).
19561983DefenseExpenditures
fromUnitedNations,
StatisticalYearbook,Annual
Issues.Defense
Burden(1978-1988)
fromUSACDA,
WorldMilitaryExpenditures
andArmsTransfens:1989.
Table A2
DATA, THAILAND
Year
Gross
Domestic
Producta
Defense
expenditureb
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
43
44
45
48
54
5g
63
68
75
84
101
108
817
1567
1390
1421
1378
1080
1570
1609
1745
1877
2055
2437
Defense
USACDA
Growth rate
burden
defense
of real GDP
(percent) burden(percent) (percent)
1.9
3.6
3.1
3.0
2.6
1.8
2.5
2.4
2.3
2.2
2.0
2.3
6
-1.6
3.4
6.9
10.0
5.3
8.1
8.4
6.6
7,9
12.2
7.8
142
JOURNALOFPHILIPPINEDEVELOPMENT
Table A2 (continued)
Year
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
Gross
Domestic
Product,
117
129
147
153
170
222
279
303
347
403
488
559
659
760
820
910
Defense
expenditursb
2990
3638
4898
5383
5721
5950
7104
7870
9987
12566
17367
22978
27019
29143
33652
34944
Defense
USACDA
Growth rate
burden
defense
of real GDP
(percent) burden (percent) (percent)
2.6
2.8
3.3
3.5
3.4
2.7
2.5
2.6
2.9
3.1
3.6
4.1
4.1
3.8
4.1
3.8
3.6
4.2
4.1
3.9
4.2
3.9
3.9
4.4
3.9
3.5
3.1
8,5
7,9
10,5
5.0
4.1
9.9
4.4
4.8
9.4
9.9
10.4
5.3
4.8
6.3
4.1
7.3
7.1
3.5
4.9
9.5
13.2
a. Billions
of baht.
b. Millionsof baht,
Sources:GDPandGrowthRateofRealGDP:IMF,InternationalFinancialStatistics
Yearbooks:
1984(1956-1959),Yearbook:1990(1960-1980),Yearbook:
1991(1981-1988).
1956-1983DefenseExpenditures
fromUnitedNations,StatisticelYearbook,AnnualIssues.
DefenseBurden1978-1988fromUSACDA,WorldMilitaryExpenditures
andArrnsTransfers,
198_
FREDERIKSEN:
GROWTHAND DEFENSESPENDING
143
Table A3
DATA, SOUTH KOREA
Year
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
t 987
1988
J/
Gross
Domestic
Producta
Defense
expenditurea
113
150
196
203
216
243
291
352
500
711
798
1024
1259
1630
2130
2724
3379
4170
5416
7569
10224
1_996
18074
24327
31323
38041
47482
54443
63833
6
11
11
13
14
15
17
21
20
25
30
41
50
65
84
101
136
171
181
254
194
771
1008
1438
1597
2252
2831
3163
3405
Defense
USACDA
Growthrate
burden
defense
of real GDP
(percent) burden(percent) (percent)
5.3
7.3
5.6
6.4
6.6
6.2
5.8
6.0
4.0
3.5
3.8
4.0
4.0
4.0
3.9
3.7
4.0
4.1
3.3
3.4
1.9
5.5
5,6
5.9
5.1
5.9
6.0
5.8
5.3
5.1
5.1
4.8
4.4
4.3
_.
l_,
4,5
-1.3
7.6
5.5
3.9
1.2
5.8
2.1
9.1
9.7
5,7
t 2.2
5.9
11.3
13.8
8.8
9.2
5.9
14.4
7.9
6.5
13.2
10.9
9.7
7.4
-2.0
6.7
7.3
11.8
9.4
6.9
12.4
12.0
11.5
,,
a.Billionsof won
Sources:
GDPend GrowthRatsof RealGDP:IMF,InternationalFinancialStatistics
Yearbooks:1984 (1955-1959),Yearbook:1990 (1960-1980)Yearbook:1991 (1981.1988).
DefenseExpenditures
from UnitedNations,StatisticalYearbook,AnnualIssues.Defense
Burden(1984-1988)
fromUSACDA,
WorldMllita_/Expenditures
and ArmsTransfers:1989.
144
JOURNAL OF PHILIPPINE DEVELOPMENT
Table A4
DATA, MALAYSIA
,,A/Ill
Year
1961
1962
1963
1964
1965
1966
1967
1968
196g
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
Gross
Domestic
Product a
Defense
expenditurea
6696
7056
7515
8056
8837
9394
9774
10160
11629
12155
12955
14220
18723
22858
22332
28085
32340
37886
46424
53308
57613
62579
69941
79550
77547
108.8
127.4
154.9
216.5
303.1
379.5
366.6
379.3
329.6
436.7
546.0
707.6
725.3
954.7
1053.8
1117.2
1324.0
1406.0
1704.0
2253.0
3332.0
3694.0
3489.0
2626.4
1850.2
Defense
burden
(percent)
USACDA
defense
burden
(percent)
1.6
1.8
2.1
2.7
3.4
4.0
3.8
3.7
2.8
3.6
4.2
5.0
3.9
4.2
4.7
4.0
4.1
3.7
3.7
4.2
5.8
5.9
5.0
3.3
2.4
4.2
4.5
2.8
Growth
rate of
real GDP
(percent)
-2.1
5.4
3.3
7.2
8.6
6.3
6.1
3.9
15.6
2.4
7.1
9.4
11.7
6.3
0.8
11.6
7.8
6.7
9.3
7.4
6.9
5.9
6.3
7.8
-1.0
1.0
5.4
-0.1
Consumer
Price
Index
97
97
100
100
101
101
99
99
98
100
102
a. Millionsofringgit
Note:Only 1960-88data used. RealGDPfor earlieryearsunavailable,
Sources:GDP (1960-1970)and ConsumerPrice Index (1961-1970)from IMF, InfematlonalFinancialStatistics Yearbook: 1990.DefenseExpenditures1961*68fromUnitedNations,
Statistical Yearbook,Annualissues.DefenseExpenditures1969-1985fromDefence Spending
in Southeast Asia, ed. ChinKin Wah (Singapore:Instituteof SoutheastAsianStudies, 1987),
p.174. Defense Burden 1986-1988 from USACDA,World Military Expenditures and Arms
Transfers, 1989.GrowthRateofRealGDPfor 1961-1970computedbytakingcurrentGDPand
adjustingusingConsumer PriceIndex;for 1971-1988from International Financial Statistics
Yearbook: 1990.
GROWTHAND DEFENSESPENDING
145
Table A5
DATA, SINGAPORE
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
USACDA
defense
burden (percent)
2._1
2.8
6:0
8.8
7.8
6.0
5.2
4.7
5.0
5.5
6.2
5.3
4.9
5.2
5.3
5.1
4.2
5.2
5.9
5.5
5.1
5.3
Growth rate
of real GDP
(percent)
i 1.8
13.9
13.7
13.7
12.5
13.4
1.5
6.3
4.1
7.5
7.8
8.6
9.3
9.7
9.6
6.9
8.2
8.3
-1.7
2.0
9.4
11.1
""
DefenseBurdenfromUSACDA,
WorldMilitaryExpenditures
and ArmsTransissues.GrowthRateof RealGDP,1967-1980fromIMF,InternationalFinancial
Yearbooks:1984;1981-1988
fromYearbook:1991.
146
JOURNAL OF PHILIPPINE
DEVELOPMENT
Table A6
DATA, INDONESIA
Year
Defense
burden
(percent)
Growthrate
of real GDP
(percent)
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
6.3
4.6
2.9
2.0
2.2
1.2
2.6
3.1
3.2
3.3
3.5
3.4
2,9
2.9
3.8
3.5
3.3
3.3
3.3
3.1
3.2
3.2
2.7
2.6
2.4
2.5
2.1
1.8
5.1
2.4
-2.4
5.5
0.0
2.3
2.3
11.1
6.0
7.5
7.0
9.4
11.3
7.6
5.0
6.9
8.8
7.8
6.3
9.9
7.9
2_2
4.2
7.0
2.5
5.9
4.1
6.5
Souroes:
DefenseBurden(1961-1966):
Stockholm
International
PeaceResearch
Institute,SIRPIYearbook1990,WorldArmamentsand Disarmaments,(OxfordUniversity
Press,
1990)Table5A.3.1967-1988fromUSACDA,
WorldMilitaryExpenditurss
andArmsTransfers,
annualssues,GrowthRateofRealGDP(1961-1980)
fromIMF,IntemationalFinancialStatistics
Yearbooks:1984;1981-1988fromYearbook:1991.
FFIEDERIKSEN: GROWTH
AND DEFENSE SPENDING
147
REFERENCES
Benoit, Emile. "Growth and Defense in Developing Countries." Economic
Development and Cultural Change (1978): 271-80.
Deger, Saadet. Military Expenditures in Third World Countries: The Economic Effects. London: Routledge and Kegan Paul, 1986.
Frederiksen, Peter. "The Relationship between Defense Spending and
Economic Growth: Some Evidence for Indonesia, 1964-85." ContemporarySoutheastAsia
(1989): 375-84.
Frederiksen, Peter and C.J. LaCivita. "Defense Spending and Economic
Growth: Time-Series Evidence on Causality for the Philippines, 195682." JournalofPhi/ippine Development (1987): 354-60.
Hsiao, C. "Autoregressive Modelling and Money-Income Causality Detection." Joumal of Monetary Economics (1981 ): 85-106.
Institute of Southeast Asian Studies. Defense Spending in SoutheastAsia,
edited by Chin Kin Wah. Singapore: The Institute of Southeast Asian
Studies, 1987.
International Monetary Fund. Internationa/Financia/Statistics
Yearbook,
various issues. Washington, D.C.: International Monetary Fund.
Joerding, Wayne. "Economic Growth and Defense Spending: Granger
Causality." Journal of Development Economics (1986): 35-40.
LaCivita, C.J., and P.C. Frederiksen. "Defense Spending and Economic
Growth: An Alternative Approach to the Causality Issue." Journa/of
Development Economics (1991 forthcoming).
Lindgren, Goran. Bibliography on the Economics of Armaments. Uppsala,
Sweden: The University of Uppsala, 1988.
Looney, Robert E., and P,C. Frederiksen. "The Economic Determinants of
Military Expenditures in Selected East Asian Countries." ContemporarySoutheastAsia
(1990): 265-77.
Stockholm International Peace Research Institute. SIPRI Yearbook World
Armaments and Disarmament, various issues. Oxford: Oxford University Press.
United Nations. Un/tedNations Statistical Yearbook, various issues. New
York: United Nations.
United States Arms Control and Disarmament Agency. World Military
Expenditures andArms Transfers, various issues. Washington, D.C.:
US Arms Control and Disarmament Agency.