7
Rsarch
Policy
WORKING
1
PAU
PubilcEconomics
Policy ResearchDepartment
Tha WorldBank
Septernber1993
WPS 1196
CorporateTax Structure
and Production
Jeffrey Bernstein
and
Anwar Shah
Investment tax credits, investment allowances, and accelerated
capital consumption allowances are more cost-effective in promoting investment than more general tax incentives such as
corporate tax rate reductions.
Policy Research Working Papes dissinate the rindings of work in pmgress and encournge the exchange of ideas arnong Bank saff nd
all othersintersted in development issues Thesepapers, distributed hy the ResearchAdvisory Staff, carry thenamnesof the authors,reflect
only theirviews, andshouldbe used andcited accordingly.The findings, interprettions, andconclu ions aretheauthon'own. They should
nd be attributedto the World Dank,its Oard of Directors,its management,or anyof iu member countries.
PolicyRarwoh
PubicEocnomics
WPS1196
This paper-a product of the Public Economics Division, Policy Research Department-is part of a larger
effort in the department to evaluate public policies for private sector development in developing countries.
The study was funded by the Bank's Research Support Budget under research project "An Evaluation of
Tax Incentives for Industrial and Technological Developmnent"(RPO 675-10). Copies of the paper are
available free from the World Bank, 1818 !4 Street NW, Washington, DC 20433. Please contact Carlina
Jones, room NIO-063, extension 37699 (September 1993, 61 pages).
Bernstein and Shah provide an empirical framework for assessing the effects of tax policy on an
array of producer decisions about output supplies
and input demands in Mexico, Pakistan, and
Turkey. They specify and estimate a dynamic
production structure model with imperfect
competition for selected industries in these
countries.
The model results suggest that tax policy
affected production and invcstinent and further
that selective tax incentives such as investmcnt
tax credits, investment allowances, and accelerated capital consumption (depreciation) allowances are more cost-cffective at promoting
investment than morc general tax incentives such
as corporate tax ratc reductions. The long-run
cost-effectiveness of these incentives - except
corporate tax rate reductions, which proved costineffective in all cases -- varies by country. In
Turkey, investment allowances and capital
consumption allowances were cost-effective. In
Mexico, neither investment tax credits nor
accelerated capital consumption allowances were
cost-effective. In contrast, in Pakistan, both
investment tax credits and accelerated capital
consumption allowances were cost-effective. In
the intermediate run, defined as tax policy
impact after one year, only the investmnent
allowances and arxeicra!zd capital consumption
allowances available to Turkish industries
proved cost-effective.
To make selective tax incentives more
effective, investment tax credits must be refundable and carrying forward investment and
depreciation allowances must be permitted. If
stimulating investment expenditure is the sole
objective of tax policy, reducing the corporate
tax rate is not a cost-effective instrument to
achieve this objective.
The PolicyResearchWorkingPaperSericsdisseminatesthcfindingsof workunderway in theBank.An objectiveof the series
is to get these findingsout quickly, even if presentationsare less than fuillypolished. The findings.interpretations,and
conclusionsin these papersdo not necessarilyreprcsentofficialBank policy.
Producedby the Policy ResearchDisseminationCenter
CORPORATETAX STRUCTUREAND PRODUCTION
by
JeffreyBernstein*
AnwarShah
lAble gf Contents
1.0
1.1
1.2
1.3
Introduction ..
.
..
..
..
....................
....
. . . .2.....
Cost of CapitalExpenditures.
Tax Structure
and Production:a DynamicTheoreticalModel
Estimation Model an6 Tax Elasticities .......................
.15
Short-Run ...........
......
Inteminediate-Run
..
Long-Run
..
Rental Rate Elasticities with Respect to Tax Instruments .
1
...
.2
.....
.19
.
.
20
21
22
1.4 Tax Policy, Impact on Investmentand Government Revenues .24
1.5Mexico
.
..........
26
1.5.2 Tax Policy Effects on the Rental Rate and Capital .30
1.5.3 Tax Incentives, Investment Impacts and Foregone
Reveues
.33
1.6 Pakistan .35
1.6.1 Tax History .35
1.6.2 Tax Policy Effects on the Rental Rate :md Capital .36
*
Of Carleton University and National Bureau of Economic Research, and World Bank,
respectively. This is one of a series of papers prepared for the research project (RPO
No. 675-10) on tax incentivesdirected by Anwar Shah, PRDPE. Research assistance for
this paper by Bjom Larsen, John Baffes, and CostasChristou is gratefully acknowledged.
The authors are grateful to Frank Lysy for comments.
-1-
-21.7 Turkey .........
1.7.1 Tax History .
43
.
. . ............................
43
..................................
1.7.2 Tax PolicyEffectson the Rental Rate and Capital ..... . . . . .
1.8 Summary and Conclusiors
............
..
..
.........
..
On the Elasticityof RentalRate of CapitalwithRespect
.
to Tax Instruments...........
On the Tax Sensidvity of Capital Stock ..........
On t*heBenefit-CostRatios ..........................
.
52
52
. . . . . .53
53
57
Endnotes ................................................
References .....
. 47
.9........................................
. .i
_.cg
I I -
1.0 INTRODUCTION
Fiscal incentivesfor investmentpromotionare prevalentin most developingcountries.
The effectivenessof these instrumentsin meetingstatedpolicy goals is an importantarea of
publicpolicyconcernyet rigorousdevelopingcountryempiricalevidenceto guidepolicyin this
as is almosteomp.etelylacldng. To addresstheseconcerns,in the past, policy makersrelied
on opinion survey of firms (see for e&.ample,Guisingerand Associates, 1985), and more
rceently,on marginaleffectivetax rate analysis(see for exampleBoadwayand Shah, 1992).
However,noneof these approachesis able to analyzethe effects of tax policy changeson the
strtvcre of production and the rate of capital accumulation.
Thispaper developsand estimatesa dynamicmodelof productionto examinetax effects
on an array of produc+iondecisionsregardinginputs and ov;tputfor six industriesin three
developingcountriesnarnelyMexico,Pakistanand Turkey.The paper evaluatesinvestmenttax
credits, investmenttax allowances,capital cost allowancesand corporate income taxes as
instrumentsfor investmentpromotion.Underan investmenttax creditcorporationsare allowed
to deduct againsttheir tax liabilitiesa fraction of expenditureson new additions to physical
capital stock. Tax credits providea direct subsidy to such activities. An investmenttax
allowance allows a deduction from taxable income based on
a
fraction of investment
expenditures.Capitalcostallowancespermit depreciationfor tax purposesas a deductionfrom
taxableincome. Corporateincometax reductionspermit a lower rate of taxationon corporate
income. The paper is organized into the following sections. Section 1.1 presents i!lustrative
calculationson the post-tax cost of capital expenditures under altemate tax policy provisions and
a historyof tax changesin threecountries. Section1.2 presentsthe theoreticalmodel. Section
1.3 specifiesthe empiricalframeworkand derives relevantelasticityformulate. Section 1.4
discusmsthe impactof taxpolicyon investmentand governmentrevenues. Sections1.5 through
1.7 prest the empiricalresults for selectedindustriesin thl samplecountries. A finalsection
summarizesthesmresults.
1.1 COST OF CAPITAL EXPENDITURES
Four tax instrumentsthat affectthe purchasepricesof capitalstocksare consideredhere
namely;the corporateincometax rate, the alloweddepreciationrate, investmentallowanceand
investmenttax creditrate. To see the effectsof tax policyon the after tax or post tax pu-,hase
prices, consider a machine that has a price of one unit denominatedin the local currency.
Dealingfirst with the alloweddepreciationrate, supposethat depreciationoccurs at an annual
rate of 30%. In addition,the expenditureon the machinemust be capitalizedand assumethat
the future depreciationdeductionsare discountedat the rate of 15%. The present value of
depreciationdeductionsbasedon declinir.,balancedepreciationis, z = d(t + r)/(r + d), where
d is the allowed depreciationrate and r is the discountrate. Thus the tax deductiondue to
depreciationis 0.77.
Next considerthe corporateincometax rte. In the presentexamplethe tax reduction
due to depreciationequals0.77u,, where u is the corporateincometax rate and the post tax
cost of the unit value of the machineis 1 - 0.77u,. If the corporateincon.S tax rate is 0.46, and
thereis taxableincome,thenthe post tax cost is 0.65 and the tax reductionis 0.35 on a machine
of unit value in the local currency.
3It is of interest to compare the tax reductiondue to depreciationdeductionsand the
reductiondue to the immediatewrite-off of the machine. In the latter case, assumingthere is
taxableincome,the tax reductionis u, Lndthe post tax cost is 1 - u,. Hencewith a corpora^e
tax ate of 0.46 the post tax cost is 0.54 and the tax reductionis 0.46. The tax reductionin the
depreiaon deductionscae is 23% smallerthan the tax reductionfrom immediatewrite-off.
Neot, considerthe investmenttax credit. Let the credit rate be v. The tax reduction
on the unit valueof the machineis zg,(l -v) + . There are three aspects to this tax reduction.
The first, is zumwhichis the depreciationpart. The secondis *zu v whichis the amountthat
the tax creditreducesthe depreciationbase. The third is v whichis the investmenttax credit.
Thusthe post tax cost of the unitvalue machineis 1-[zU,(l-v) +v]. If u. is 0.46 and v is 0.10
then the tax reductionis 0.42 and the post tax cost of the mwchineis 0.68.
Some countries, for example Turkey, rather than offring a credit for investment
expenditureallow a fractionof these expendituresto be deductedfrom taxable incomein the
year such outlaysare made. This is an investmenttax allowance. Under such a regime, the
post tax cost of the unit valueof machineis 1 - [zu1 + u0O]. If z = 0.77, u
=
0.46 and # (the
allowancerate) - 0.10 then the tax allowancecontributes0.40 to ax reductionwith the final
cost of the machineequal to 0.60.
Table 1.1 showsexamplesof the post tax cost of unit value machineryand equipment
for three countries;Mexico,Pakdstanand Turkey. The highestpost tax cost of a unit value of
-4Table 1.1
Cost of a unit value of capitalexpenditure
Thscont
Mexico'
0.46
0.53
Paklstanb
0.43
0.52
Turkeyr
0.46
0.53
Cu,= 0.42, straightlinedepreciationat 0.10, this is an averagerate, v = 0.30,
and z = 0.811 for r = 0.O5and z
=
0.S77for r
=
0.15.
-u = 0.55, this includesthe supertaxrate, declning balancedepreciationat 0.10, v
= 0.30, and z = 0.7 for r =0.OS and z = 0.46 for r = J. 15.
'u, = 0.46, decliningbalancedepreciationat 0.25, investmentallowancerate -
0.30, this is the minimumrate allowedand z - 0.875 for r - 0.05 and z0.719 for r
O.
0 15.
-5.
capital expenditure is found in Turkey, followed by Mexico and I-aldstan. As fuure
depreciationdeductionsare discountedat a hignerrate. their value diminishesand the post tax
cost of the expenditurerises. This can be seenf:oni the table, as the secondcolumnfiguresare
higherthan those foundin the first column.
1.2 TAX STRUCTUREAND PRODUCTION:A DYNAMICTHEORETICALMODEL
The technologyof a representativefirm withinan Mndustry can be definedas
y, X f(Kv,.,v,, A14, Aj
(1)
where y is the outputquantity,K is the m dimensionalvectorof quasi-fixedfactors,v is the n
dimensionalvector of variablefactorsand A is the indicatorof the level of technology. The
productionfunctionis denotedby f, which is definedfor nonnegativeinput quantities,and is
nonnegativewithpositivemarginalproducts. The productionfunctionalso decline.withrespect
to the net investment vector, AK = K,- K,. . Adjustmentcosts are represented through the net
investmentvectorin the productionfunctionand are measuredas foregoneoutput. The cost of
changinga quasi-fixedfac:oris the loss in outputthat could have been produced. Adjustment
costsare, thereby,internalto the productionprocess(see for example,Treadway[1971,19/4],
Mortensen(1973]and EIpstein(1981]). The subscriptt representsthe time period I
Quasi-fixedfactors are also referred to as capitalinputs. In this model, capital inputs
relate to various ypes of plant and equipment. The stocks of the capital inputs accumulate
accordingto
IK =
+ (,a- 5)Kv.,,
(2)
whereI is the m dimensionalInvestmentvector, a is an m dimensionaldiagonalmatrixof fixed
depreciationrates such that Og
s
s 1,
I - 1, ..., m. It is assumedthat capital servicesare
proportionalto the capital stocks (see Bernsteinand Nadiri [1988D).lIn addition, I. is the
dimensionalidentitymatrix.
Finns sel their products,hire or purchasefactorsof producuon,investin captal stocks
and financetheir operations,such that the flowof fundsis givenby,
p y, - wsTv_t-Ig + AB, +1P ,&ANst
- r,B, I - TOt- D t -0.
(3)
The productprice is denotadby p, w is the vector of variableinput prices, q is the vector of
capitalpurchaseprices, B is the value of outstandingbond issues,AABB, - Bj. is the value
of net bondissues (netof retirements),p, is the price of shares,N, is the quantityof ouuwanding
shares, AN,, - N, - N., are new share issues, rbis the interestrate on bonds, T. aie income
taxes and D is the value of dividends.
The flow of funds can be further decomposedby consideringincome taxes. First,
investmentincentivesare often in the form of credits sucihthat at time t with a credit rate of
O<V,zt,<-j,...m. the ith capitalstock investmenttax credit is,
rX,,-V,,qw,I,
iZ,..,
(4)
Second,there are capitalcost allowancesassociatedwith the depreciationof the capital
stocks. In general,depreciatioz'deductionsequal di, on a unit value of the originalcost of the
ith capitalstock of age r. Sincecapitalmustbe fully depreciated,then it must be the case that
E
1.0
1,
a- 1, ..., m. The capitalcost allowanceat timet for the ith capitalstockinstalled
at differenttimes is,
-7CCA1 9 * s qs1 , I,.,(l -,v1,)d,r
'
...
(5)
m
O0
where 0 5 #, S 1 is the prportion of the investmenttat credit v.hih reducesthe depreciztiori
base for tax purposes.
with sect to the hbor inputsthere are no payrolltaxes -it with respectto the
MhMd,
intemediateinputsther are valueaddedtaxes. Let the tax rate on hejth variableinputat time
t be n <q, < 1 and so the post value added tax is 2;. ogw,.
The incometax is definedat
ime t by therate O < u,, < 1, basedon revenuenet of the post value added tax cos+ ,f'the
.able
inputs,net of inteet payments,net of capitalcost allowancesznd net of investment
tax creditsl. Thus incometaxesat time t are
X_
uOI(P.We
T
%-)V-rk,
s-iCW-iUC
+ wTt v
(6)
whereI, is the n dimensionalidentitymatrix, (sois a diagonalmatrix of value added tax rates
(includingpayroll tax rates), i is the m dimensionalidentity vector, CCA and ITC are m
dimensionalvectors of capital cost allowances and investment tax credits respectively.
Substitutingequation(6) (the incometax equation)into the flow of funds equation(given as
equation(3), yields,
,Yty- wT as - p),vj(I - %)
* [D/(pl..)
A (PN)
q;- I +i,,(u,,CC-,k+lTC,
+ Apr/Pg
1 PgNg-1 + rbt(I - u,)BD,X
- ABt
(7)
8The leP. side of equation (7) shows revenuenet of tax, net of variableinput cost and tnetof
investmentexpenditures.The right side of the equationshowsthe flowof fundsto bondholders
and shareholders. Equations(1), (2). and (7) summarizethe technology,capitalaccumulation
and flowof funds for the representativefirm in the industry.
Turningto the natureof marketstructure,the first marketto be consideredis the product
market. P..oductdemandis representedby
p,UD (Yr,e,),
where Y M.
(8)
J is industryoutput, with"hesuperscriptrepres nting the particularfirm, and
c is a vector of exogenousvariablesaffectingproductdemand. The iiverse prod' :t demand
functionis given by D, which is defined for nonnegativeindustryoutput, nonnegativeand
decreasingin industryoutput. Impliedby the inverseproductdemandfunction,is that firms
withinan industryproduce homogeneou,products. Moreover,dependingon the conjectural
relationshipbetweenthe outputof a firm and industry,the productmarketin the modelcan be
competitive,monopolisticor oligopolistic(see Bernsteinand Mohnen(1991]).
Second,the variableand capitalinputmarketsare assumedto be competitive.Thusfirms
face exogenousvariableand quasi-fixedinputprices. The last set of marketsare the financial
markets. Giventhe less developednatureof the economy,firms are not able to affecithe rates
of return on their shares or bonds.These rtes of returns are essentiallyconstrainedby world
financialmarkets. Definefinancialcapitalas V, = P,N, + B, and so AV, = A(p&,
1 ) + AB
1,
then equation(7) can be rewrittenas
Fe- [r,, + r,(l - u, ) A,_J](I+ 1_,)-' V,_1- AV,
(9)
9-
whereF, is the left sideof equation(7), whichis net aftertax revenue,the rate of return on the
shares of a firm isr rS.
5 ,
+±
, and the leverage ratio is A, 1 -
. The rate of
return on sharesconsistsof the payoutratio, whichis dividendsper value of outstandingshares,
plus the capitalgains (or losses)on the shareprices. The leverageof a firm is A whichis the
ratio of debt to financialcapital. Definep as the coefficientof V, in equation(9). It is the
rate of return on financialcapitalwhichis a weightedaverageof the rates of return on equity
and debt. It is assumed that the rte of return on financialcapital issued by a firm is
exogenous.
the objectiveof a firm is to operate in the interest of its owners by maximizingthe
expectedpresentvalue of the flow of funds to its shareholders. In the contextof the present
model,becausethe ratesof returnon equityand debtcapitalare exogenous,and thereforecannot
be influencedby shareholders,the objectiveis equivalentto maximizingthe expectedpresent
value of the flow of funds to shareholders and bondholders. In other words a firm maximizes
the expectedpresentvalueof financialcapital. The objectivecan be obtainedfromequation(9).
Solvingfor V1 and applyingthe conditionalexpectationsoperatoryields.
J, -, E
i a(t, ) P,y,[Ps- W,J,
-Qr I, - " M,J
(10)
where EAis the expectationsoperatorconditionalon 'nformnation
knownat time t, the discount
rate
is therate ofreturn on fancial capital,a(t, t) -1, a(t, t+l) a (1 + p)-1 , p mp(i- u) is the
W,- wt (1 -u; ) (1 - j ) j - 1,...n are theafter tax variable factor
after
taxproduct price,
prices, and Q is an m dimensionalvector of after tax capital stockpurchaseprices,
-
Ql, - qi, (1 - V.' - i
10-
a( t,s + r9)a( t,s)l ue,,.,r(1-I,
d)
s.0
and M is an m dimensionalvector of tax reductionsat time t, due to capitalcost allowances
arising from past investment expenditures, M,
-
u,5 i
q
, I,
(1 - *1 1 .,v,)d 1 ,* At any
time t, M does not affect outputsupplyand inputdemanddecisionsbecausefrom the vantage
point of the present the vector is predetermined. A sgnificant featureof a dynamicmodelis
that current and future talxrates, cedits and allowancesare explicitlyaccountedfor in the
analysis.Indeed,the future taxpurchaseprices of the capitalstocksshowsthe array of current
and future tax policyinstrumentswhichaffectthe analysis.1
A firm maximizesthe expectedpresent value of the flow of funds (in other words the
right side of equation(10))by selectingoutputsupply, variableinput and investmentdemand
subjectto the productionfunction(equation(1)), capitalaccumulationequations(equation(2)),
the inverseproductdemandfunction(equation(8)), the exogenouscurrentand future after tax
factor prices and discountrates. This program can be solved in two stages. The first stage
relates to the short run decisionsand the secondstage concernsthe intertemporalproduction
choices. Conditionalon the capital stocks, output supply and variable factor demand are
determined. With this solution,a firm then proceedsto determinethe demandfor the capital
inputs. In breaking the prob.em into two subsets, the first stage solution or short-run
equilibriumis foundby maximizingafter tax variableprofit at eachpoint in dme. Thus
ma*x P,y, - W* v.
( y,V
)
(11
sub3ectto equations(1) and (8), and giventhe capitalstocks. Substitutingequation(8) into(11),
the first order conditionsare,
D(Y"p~~
-W, + lf,
where t
v,
a Y(oD/M/p
~
~~~e
l
O
(12.2)
is the inverseprice elasticityof productdemand,e . (py/a)yyy is the
conjecturalelasticity,X is the Lagrangianmultiplierand the superscripte denotesequilibrium
values.1 From equation (12.1), in short-runequilibriuma firm equates after tax marginal
revenueto marginalcost. The Lagrangianmultiplierequalsmarginalcost. Equationsc: (12.2)
implies that relative after tax variable factor prices equal relative marginalproducts of the
respectivevariablefactors. Equationset (12) holds for all time periods and, of course, for all
firms in the industry. Equationset (12)showshow tax policyaffectsthe short-runequilibrium
The corporateincometax rate doesnot directlyaffectthe short-runequilibrium.From equation
set (12) relativevariablefactorprices and the relativeproductprice (all prices are normalized
for exampleby the nth variablefactorprice)are independentof the corporateincometax rate.
The reason is that the corporateincometax is a tax on variableprofit in the short run, and as
a consequence,it is basedon the residualof the short run incomestream. Third, the corporat
incometax rate, like the investmenttax credit and capitalcost allowancerates indirectlyaffect
the short-runequilibriumthroughtheirinfluenceon the demandfor the capitalinputs. Change
in these rates affect the after tax purchaseprice of the capital inputs and thereby alter th
demandfor the quasi-fixedfactors. These changesin the capitalinputlevelsthen influenceth
shortrun supplyof outputand demandfor the variablefactorsof production.
The short-runequilibriumconditionsare consistentwith a numberof product market
structures. The conjectual elasticity, e, shows the nature of firm interdependencein the
* 12-
product market. If e - 0 then the productmarketis purely competitiveas firms are price
takers. If o
-
1 then the product marketis purely monopolisticas there is only a single
producer.If e * y/y then the productmarketis oligopolisticand the firms are characterzedas
Cournot-Nasholigopolists.In the latter case, if firms have the samemarginalcost in short-run
equilibriumthen from equation(12.1), firms have the same conjecturalelasticityin short-run
equilibrium.
An alternativeway that the short run equilibriumconditionscan be characterized,
emphasizesboth product market imperfectionsand the dual relationshipbetweenprice and
quantityeffectson variableprofit. Considera first order approximationto the revenueof a firm
in equilibrium,
D( Y,, c,)y - D(Y', eg)yt D(Y,',et) (I + 4 Vj,](y, -y,')
(13)
Collectingterms yields,
pty - pt,(l + etE8)yt
- p,*y&','e''
(14)
From equation(14), total revenueequalsrevenueearnedin a purelycompetitiveproductmarket
plus the additionalrevenue eaned in equilibriumbecauseof oligopolypower.2 Definingthe
purelycompetitiveor shadowproductprice as p,' . pt'(l
+ g,)e,) and the after tax shadow
productprice as P,'- pt'( - u.) then the short-runequilibriumconditions(equationset (12)can
be obtainedby
max.
P:y, - W. v.(
(yy, v,)
- 13 -
subjectto the productionfunctionand the givenlevelsof the capitalinputs. Thus firms act as
if they, maximizeaftertax shadowvariableprofit, whichis definedas aftertax shadowrevenue
minus after tax variable input cost.1
The reason is that the degree of product market
imperfectionis capturedin the definitionof shadowproductprice.
The short-runequilibriumconditionscan be substitutedinto (15) to obtainthe after tax
shadowvariableprofit function.
-(Pe,
I
W,,K,
K
AK,,A,)
(16)
where ?e is after tax shadowvariableprofit,III is the after tax shadowvariableprofit function
whichis definedfor non-negativeafter tax prices and capitalinputs,increasingin the after tax
shadowproductprice and capital inputs,decreasingin the after tax variable factorprices and
net investmentin the capitalstocks.2
The dual relationshipbetweenprice and quantityeffectsin equilibriumcan be seen by
differentiatngthe aftertax shadowvariableprofitfunctionby the after tax shadowproductprice
and the after tax variablefactorprices. This yields,
y5 ' -4
(17.1)
v - -V l3n
(17.2)
The short-runequilibriumoutputsupplyand variablefactordemandscal be obtainedfrom the
after taxshadow variable profit function. It implies that short-run equilibrium can be
characterizedby equations(16), (17.1)and (17.2). The attractivefeatureof thisapproachis that
reducedform outputsupplyand variablefactordemandequationsare readily obtainablefrom
the after tax shadowvariableprofit function.
- 14The secondstage of the programinvolvesthe deterninationof demand for the capital
inputs. This stage relates to the intertemporalaspects of productiondecisions.Capital input
demandcan be obtainedby consideringthe expectedpresentvalue of the after tax shadowflow
of funds.The objectiveis to
max.~
t...(8
subjectto the capital accumulationequations(denotedby equationset (2)). The first order
conditionsfor this problemat any timeperiod, after substitutingequationset (2) into (18), are,
V(ft'1aAK,)
- Q, +Ea (s,s+l)[v(8x:.
a)
1 I8Kg)
-[V(8iax
1 /OAK
(19)
+o1 -a) Qff1J -On
Equationset (19) implies that the marginalcost of a capital input is equatedto the expected
marginalbenefit of that capital input.19 The marginalcost consistsof two components;the
after tax marginaladjustmentcost and the after tax purchaseprice. The expectedmarginal
benefitconsistsof three components;the expectedafter tax marginalprofit, the expectedafter
tax adjustmentcost savingand aftertax purchaseprice savingfrominstallingand purchasing(or
renting) the respectivecapital input in the previous period. Equation set (19) shows the
intertemporaltrade-offbetweengreaterexpectedfuture after tax profit due to increasesin the
capitalinputsand smallercurrentafter tax profit resultingfrom increasesin the capital inputs.
It is importantto notice that this equationset containsall current, as well as, expectedfuture
tax, credit and allowancerates. These rates enter throughthe after tax purchaseprices of the
capital inputs.
15 -
The completeset of equilibriumconditionsare given by equations(16), (17) and (19).
Equations(16) and (17) define a short run equilibrium,while equations(16), (17) and (19)
define a temporaryequilibriumof producer behavior. In the temporary equilibriumoutput
supply,variablefactorand capitalinputdemandsare determined.
1.3 ESTIMATIONMODEL ANDTAX ELASTICMTIES
This sectionparameterizesthe dynamicmodel of productionpresentedin section 1.2.
Thedynamicnatureof the modeloffersmanyadvantagesin determiningthe impactof taxpolicy
on outputsupplyand inputdemands. First, the modeltreats capitalinputsdifferentfrom other
factorsof productionas producersmustincuradjustmentcoststo investin capital. Second,the
modelallows for short-run,intermediate-runand long-runeffects of tax policy initiativesto
differ. These effectsdiffer accordingto the extent that capitaladjustmenthas occurred.
In the empiricalspecificationof the model, it is assumedthat there is one output, two
variablefactors(laborand intermediateinputs),and one quasi-fiscalfactor. In order to estimate
the dynamicmodel of productionwe need to parameterizethe after tax normalizedshadow
variableprofit function(givenas equation(16) ). This functionis assumedto be normalized
quadraticand is writtenas
+0-5[P
A
Ps'+POO'](0
+PgWk'+ Paff 8-1
+ pPWgS + pop+
P&W&,
1
+
'As
+ PkW&A + Pk3Kh$ + O.SIAsK2
~~~~~~~~~~
-
16 -
where s, is the after tax normalizedshadowvariableprofit (normalizationis by the after tax
price of intermediateinputs), PI, is the 'fter tax normalizedshadow price of output (see
equations(13)-(1S)),W,. is the normalizedlabor inputprice or normaliz wage rate, K, is the
capitalinput, A, is the indicatorof technologyand &K,representsnet investment.All variables
are indexedby the timeperiod s.
From the profit function, we find the equilibriumconditionsfor output supply and
variable factor demandsby differentiatingthe after tax no-mnalizedshadowvariable profit
function(e4uation20) withrespectto the relevantprices. Thus we obtainthe followingspecific
output supplyand input demandfunctions(whichwe are givenas equations(17.1) and (17.2)
in section1.2),
Y - pp + PP3 + PpWI + PPkK8+ P1 AI.
(21)
-vi - PI PX
+ W+ P
(22)
Since -v
+ PA-,
+
Pf3A,
no -P ay+W1v1, then the intermediate input demand equation is
-v.
=Po
+ Pa-I
+O.5p kK
+ %A, + ,A,1
+0.5P,A,' -
-05
3P,
-O
,P,PW
1 . + PkIK,lA,
(23)
+ 0.Sp,AK2
Thus, equations(21), (22)and (23)definethe short-runequilibriumconditionsbasedon
the normalizedquadraticafter tax shadowvariableprofit function. These equationsshow how
after tax outputand variableinputprices affectoutputsupplyand variableinputdemands,given
the levelsof the capitalstocks.
- 17 -
The equilibriumconditionfor the capitalinput, is givenby equation(19) in section1.2.
Basedon the normaized quadradc after tax shadowvariableprofit function, the equilibrium
conditionfor capitalcan be writtenas
pIAK9 - Q +(I+p)-Y[,(pk
(
+ PI&K5
+ P
+ PaA$., - ,,AK,.,
+
I1&W
(24)
+ Q,1(1 - 8))] - 0
where Q, is the normalizedafter tax purchaseprice of capital, 6 is the depreciationrate and p
is the discountrate.
If we assumethat after tax relativeprices, discountrate and technologyindicatorare
expectedto remainconstantthen we obtainthe following,
-%Ar,1.. + (1+p)pYAK$ + p+Ka+ k + p,
paWk+pw, - Q,(1+p) + Q
,(1-d)
mo (25)
Re-arrangingwe get
-p/a1.
1
+(pi (2+p)P)K,, -(I+p)PuK.. * W-s-(Pk + POP$ PRWb+
P2M)
(26
wherethe normalizedafter tax rental rate is Wk.-Q,(p+8).
Equation(26) definesa secondorder differenceequationin terms of the
capitalstock. The solutionto this equationis a flexibleaccelerator,
Ks - K2 1 - m (K, -K,-1)
(27)
whrce m a - OS(p + Pkk/P -[(P + Pkk/ g )+4Pkk pt lo-) is the speedof adjustmentof the
capital stock and the long-run capital stock is K, -
(% p+,P,P'
+ P&Wu+ P
Therefore,by combiningequations(7) and (8) we get,
- Wb) -
-A
-
+4PAJPJO"(Pk+PW
18 -
@--
(2+p
8)
+(1+0.5(p+ PMPi,-[(p + Puh
+4pidp;))X"1
for the capitalinput. It is a functionof the relativeafter
Equation(9) showethe demand
tax outputprices, variableinputprices and rentalrate, along with the discountrate and lagged
quantityof the capital input. This equationis nonlinearin the parameters.
TMeestimationmodelconsistsof the systemof equationsmadeup of equations(21),(22),
(23) and (28). These equationsdescribea temporaryequilibrium. There are four endogenous
variablesoutput supply Y,, labor and materialinput demandsv,, and v. and capital input
demand,K,. In addition,in the productionmodelthe exogenousvariablesare the normalized
after tax prices, P,W,h,WJ,, the discountrate, p, laggedcapital,Kg,,the technologyindicator,
A, . The modelis linearin the endogenousvariablesand nonlinearin the parameters.
The modelestimatesare obtainedbyjointly estimatingequations(21), (22),(23) and (28)
using the maximumLikelihoodestimator.The estimatedprofit functionmust be convex in
prices. Thus the parametersmust satisfy p.,> 0 p > 0 and p. - P
>
0. In addition,the
profit functionmust be concavein capitaland net investmentso that, Bkk
<0 and Bi< 0.
An importantfeature of this model is that there are adjustmentcosts associatedwith
capitalaccumulation.These costspreventproducersfrom immediatelyadoptingtheirlong-run
desiredlevelsof capital,and therebyalso labor, materialsand output. Producersadjust towards
-
19 -
the long run. The speedof adjustmentis givenby m in equation(27). The dynamicadjustment
processhas implicadonsfor the effectivenessof tax policychanges. For example,in the shnrtrun output supply dependson existing capital, but not on the rental rate. This means that
changesin the capitalcost allowancerate whichalter the rental rate of capitaldo not have an
effect on the supply of output. However,as capital adjustmentoccurs and the capital input
changesin responseto the new capitalcost allowancerate then outputsupplyis affectedby the
new rate. Thus, in a dynamic context it is important to distinguishbetween the short,
intermediateand long-runeffectsof tax policy. In the short run, no capiti adjustmenthas
occurred,in the intermediaterun capitaladjustmenthas occurredfor one period, and in the long
run, the capital adjustment process has been completed.
Short-Run
The short-runequilibriumconditionsare based on equations(21), (22), (23) and (28)
-Theshort-runequllibriumconditionfor outputsupplyis
+p,,P:+p$wi+p,&K,+1,A
Y-'
(29.1)
The laborand materialinputshort-rundemandfunctionsare
-
-'
.
*
p1 py +PsP:l+PziAa
Pa+Puwb
o
O
Ph%.
e,
+ 0.5p(I, 5
+,^
-*~
0.58
0.5.,NA. 2 - P,P.'WI
(29.2
_,?
0.5g
+ P1*,.IA,
-
The equationfor the short-rundemandfor the capitalinputis
(29.3
-
20 -
-0.5
PA r1
Pik11
Pi,+ lPM~
+ PpkaPs
+ PaW +
*PkA,
(t
A
it
Pi,)
-
Wx)
(29.4)
+ (1 - o- p + Phk + E(P + PAkp+ 4Pi*] )bJ
where the superscripts for the endogenousvariablessignifiesthe short-runequilibriumof the
demandfunctions.
The short-runequilibriummagnitudesof outputsupplyand inputdemandsare determined
in the followingmanner. Theshort-rundemand
for capitaldependson predeterminedvariables.
These variablesare relativeafter tax prices, indicatorof technology,discountrate and lagged
capital input. Next, the output supply and variable input demands are simultaneously
determined. These variablesdependon the aftertax relativepricesof outputand labor (not the
rentalrate), the technologyindicatorand laggedcapitalinput.
Intermediate-Run
The equationsfor the intermediate-runare derivedfrom the short-runequations. The
intermediate-runis definedwithrespectto the capitaladjustmentprocessafter one period. The
intermediate-runequilibriumconditionfor outputsupplyis
Y'.+
I
pp+
* P,P + P
+ P,,K,'
*j%
+ P>A
(30.1)
The labor and materialinputdemandfunctionsfor the intermediate-runare describedas
-VI+I
,Pi + PWAU
+ Ppf': + pkKs + Pi.Aa
(30.2)
-21-Va.-
-
PO
+ p+,-P 5
+ 0.50e,
+ 0.SPWr2
-
P,S2 - O.Sp.w"
P/W
A
+ P -
(30.3)
+ O.SpU(K8.j - K8Y
The equationfor capitalinputintermediate-rundemandis
*
(+.5IP)
[(p + §WpM2
(P + pdp
+ 4PI3Pdi0J5)(Pk+ PA'
* (1 + O.S(p+
-
+ PkW
+
P
-
WA(30.4)
(30.4)
p + p&fi)2
+ 4p,Jpj 0.5) K,'
Giventhe technologyindicatorand relativeprices, these equationsshowthe equilibrium
after one year. The superscript i indicates the intermediate-run. The intermediate-run
equilibriummagnitudesof outputsupply and input demandsare determinedin the following
manner. The intermediate-run
demandfor capitaldependson predeterminedvariables. These
variablesare relativeaftertax prices, indicatorof technology,discountrate and short-runcapital
input. Next, the cutput supply and variable input demandsare simultaneouslydetermined.
Thesevariablesdependon the after tax relativeprices of outputand labor (not therental rate),
the technologyindicatorand the short-rundemandfor capital.
Long-Run
In the long-run,AK, - 0. Thus, investmentin the long-runonlyoccurs for replacemen
purposes. The long-runoutputsupplyequationis
yS
-
+
f P^W
*P
+
P;X
+ PwA,
(31.1
-
22 -
The labor and materialinputuemandequationsfor the long-runare
I
I~ + -
p1
S
pph?6
+ PA
Po + PkKXI
-
plc
~~~~~~~~~~~~~~~~~I
- O.SPPP
-
(31.2)
OPiWs
+ 0.5sp;K + 0.5s P + O.Sp",< - pb,bW,
(31.3)
+ PjaK A,
Capitalinputdemandis givenby the followingequation
K,
-
(-'/Pi,)[Pk + pA,P;+ PI*Wl + PhA, - W,J
(31.4)
In the long-run,since the ca;.tal adjustn 'nt processis completed,output suppliesand
inputdemandsare functionsof the iong-rundemandfor capital. Thedemandfor capitaldepends
on exogenousvariables. Once this denlandis obtained,then outputsupply,labor and material
demandscan be determined. Sincethe long-rundemandfor capitalaffectsoutputsupply,labor
and materialdemandsthen the rental rate affects thesevariables.Indeed,in the long-runall
inputs are variablefactors.
Rental Rate Elasticities with Respect to Tax Instrments
In order to determinethe effectof tax policyin stimulptinginvestment,it is necessary
to determinethe tax irnstrumentelasticitiesof capitaldemandin each of the productionruns.
The tax instrumentlelasticitiesconsistof two components.The first elementis the effectof the
tax instrumenton the after tax relativerentalrate of capital(sincethis is the only relativeprice
directlyaffectedby the tax policy). The secondcomponentis the elasticityof the rental rate on
the demandfor capitalin each of the productionruns.
-
23 .
We now consider the effects of the tax instrumentson the after tax relative rental rate.
The elasticity of the after tax rental rate with rspect to the lTC rate is
-Q,(P
+ 8)(1 U(&z
u+ 5/8U))U&/,(1-u)O
(32)
Increases in the lTC rate lower the relative price of the capital input. In cases where investment
tax or incentive allowance (IIA) exist, the elasticity of rental rate of capital with respect to
allowance rate is:
ca,v -Q(p + d) (u + g(8'aI*a*))*a,/Iw*u(1
- u) < o
(33)
Next, the effects of changes in the capital cost allowance (CCA) rate also operate through the
rental rate. This elasticity is
ez,,w - -Q(p
+
8)u.,(&z,/t3dd,fW,(l - u) < 0
(34)
I.creases in the CCA rate lower the relative price of the capital input.
The corporate income tax (Cm rate affects the normalized or relative after tax rental
rate. However, the CIT does not directly affect the other relative output and input prices. The
CIT elasticity on the rental rate is
C*" -
Q(p
+
,)(1- V, -
Z)u-IWb(1 - u.)>
0U
0
(35)
Clearly, decreases in the CIT rate cause the relative price of the capital input to fall.
The effect of tax policy on capital demand in each of the short, intermediateand long-run
is obtained by calculating the tax effect on the rental rate and then multiplyingthis effect by the
rental rate elasticity of capital demand.
-24-
1.4 TAX POLICY,IMPACT ON INVESTMENT
AND GOVERNMENTREVENUES
In this sectionwe present the results of changesin each tax instrumenton the demand
for capitalper cost to the governmentof stimulatngcapitaldemand. This ratio we refer to as
the benefit-costratio.
Investmnent
Tax Credit and Allognc
For an investmenttax credit, the changein governmentrevenueis
a
Qt(14 - (l-a)K: 1 )v,
(36)
The superscripte denotesthe particularequil£brium,e = s,i,l for short, intermediateand longrun. For an allowancewitha rate of *, thenin the formula,v, is replacedby ,u . For a 1%
changein a rate multiplythe formulaby 0.01.
Cagital Cost Allowance
If depreciationfor tax purposesis decliningbalance, and tax credits do not affect
depreciationfor tax purposes,then the changein governmentrevenueis
AGR; -
-J(
(1-8)K-i)Ou
p
(37)
If depreciationfor tax purposesis straightline,and tax creditsdo not affect it, then the
changein governmentrevenueis
AGR;
- Q((v)-(1-8)R
))(
(38)
- 25 Corrate
Income Tax
The base for the incometax rate is revenuenet of variablecost, interestpaymentsand
allowances(all allowances,for examplecapitalcost and investment). Definethe base in year
s as
P
psy- -_8ysWtV
E
-
-
CCA, - IIA
(39)
where (withone type of capital, see equation(5)).
(40)
cCA* - E;.oQOs.-*d
5
where I. "-
(KR'
ITA.
-
(--8) KR' 1 ).
*, Qs (KR -
Also
(1-8)K:)
Now the changein governmentrevenuein this case is
AGR, - E:uc
Bnefit-Cost Ratio
;
=
Ac
(41)
where the numerator is the nominal value of capital (before tax, not normalized)in the
appropriateequilibrium,multipliedby the elasticityof capital with respect to the j tax
inmtrument(investmenttax credit, tax allowance,capital cost allowance,income tax). The
numeratoris the additionalcapitalgeneratedby a specifictax instrument. The denominatoris
the cost to the governmentof generatingthe additionalcapital. The ratiodenotesthe benefit-cost
ratio.
- 26 -
1.5 MEICO
1.5.1 Tax History
The structureof corporateincometaxationin Mexicohas undergonemajorchangesin
recent years. During the 80's Mexicancc-porate tax system allowed indexationof capital
consumptionallowances only. Full indexationof the corporate income tAx base is now
permitted. With indexation,corporationsare no longer allowed to deduct the inflationary
component of interest expenditures nor would they have to accumulatethe inflationary
componentof interest income (see Gil-Diaz, 1990, p. 79.) TaxableProfits (definedas gross
receipts minus purchasesand businessexpenses,and net losses carried forward from other
periods)are subjectto tax at a rate of 35% (a rate of 42% prevailedin the pre-1987period).
Depreciationdeductionsare indexed or as an alternative,the present value of depreciation
calculatedat a discount rate of 7.5% may be deducted fully in all regions except major
metropolitanareasand in all sectorsexcept the automobiles.In major metropolitanareas only
60% of suchvalue can be deductedin the first year and the remaining40% subjectedto capital
consumptionallowances.
It is instructiveto comparethe Mexicantaxationof businessincomewith a few of its
capitalexportingpartnersnamelyUnitedStatesand Canada. Table MI showsthat Mexicohas
movedsome distancetowards a cash flow type of taxationby allowinga deductionfor the
present valu'. of the schecduleddepreciationallowancesfor the life of each type of assets
calculatedat a 7.5 percent annualrate of interest(see Gil-Diaz, 1990). Tax incentivesregime
in Mexicohas also undergonesignificantchangesover time. Duringthe past two decades,tax
policywas seen as a majorvehiclefor regionaland sectoraldevelopmentwhile revenue
- 27Table NI
Nexico:
TaX regime
Corporate income t
-H-::-unerol1
Toxation of DIsinmes Income, A Coapartive
(percent)
Perspective
(1991
S--:::-v.!:-ffexi
35 * 3.9 * 38.9
tkIted Statei
34 + 6 40
Withhotding
tax rates
Interest
DivIdends
Technology trenefer fees
Royalties
3S
0-40
21
40
30
30
30
30
Indexation of deductfons
Loss carry forward
Losscarrybackiard
Niniumu/atternative
Ninima tax
Fult
5
0
2 assets tox
tat gairm toxation
Capitalgalintaxation
Coverage
Indexation
Rate
No
Mo
15
7
3
3
20X an taxableincom inclusive 0.175Yan capita(in exces of $10
of tax preference
*Illioncreditable
againt 3X
surtaxon corporate
profits
Fult
Full
35
FulL
No
34
Two-thirds
No
28
go
No. PresentValueof CCAs
lediately daductible
Yes
No
Yes
Mo
rate:
fDvidends deductien
Fullexpensing
of investumnt
Investment tax credi s
10)
19M
-
Energy irnestmnt.
rehabilitation
of real estate,
targeted job credit
O9)
28 + 15
43
26
25
25
25
Regiontl wnd R&D
2/ In Mexicotheprofit-sharing
rateand, in the UnitedStates nd tan d, the averae*
provincial
or statetax rates are addedto the basicfederalrate.
Source: Ugerte(1966),PriceWaterhouse
(1992),Internationlkureu of fiscalOocumentctien
(196), and Cil-Dlaz(1990).
-28
-
implicationsof thesepolicieswere overlooked. A brief reviewof historicalchangesin the tax
incentiveregimein Mexicois presentedbelow:
1955-1972: Between20% (for secondaryindustries)and *0% (for basic industries)
corporateincomeof Mexicanmajorityownedenterpriseswasexemptedfromcorporatetaxation
for periods varyingbetweenfive to ten years. The same industriesalso could receive, upon
application,exemptionfrom certainindirecttaxes and importduties on capitalgoodsimports.
1972-1979: Industries that were seen to promote decentralizationand regional
developmentwere grantedimport duties relief varyingfrom 50% to 100%and reductionin
corporatetax liability ranging from 10% to 40% dependingupon their locationand type of
activity.
1979-1986: The practice of import duty exemptionwas continued. In addition, tax
incentivescertificates(CEPROFIS)providingtax creditin therangeof 10-25%,dependingupon
location,and type and size of the industry, for investmentin physicalassetswere introduced.
These certificateswere negotiableand could be used againstany federaltax liabilityby the
holder. These certificatesprovedquitepopularand in 1983amountedto 0.83 percentof GDP
in revenue losses. While the manufacturingsector was a major beneficiaryof this scheme,
mining,agricultureand transportationindustres also receivedsignificantamountof resources.
Amongthe manufacturingindustries,paper and publishing,chemicalsand food and beverages
receiveda majorityof the assistance.
WhileCEPROFISwerethe mostimportantfiscalincentive,Mexicangovernmentoffered
also offeredspecialincentiveswere exportpromotion(CEDIS),developmentof duty-freezones
specialtax preferencesto automobile,cement,publishingand miningindustries.
- 29 1987-1900: The tax incentivescertificates scheme was significantlytightenedand
targetedto priorityindustriesand preferredzone. Top taxcredit rate for CEPROFIwas raised
to 40% of totalphysicalinvestmentin 1986. In additionMexican-ownedenterprisesare eIigibl
for employmenttax creditup to 30% of three times the annualarea minimumwage multipli:
by the numberof newjobs created.
Startingin 1989, full expensingof the present value of capital consumptionallowances
calculatedusing a 7.5% discountrate was offered as an alternativeoptionto standardcapit
consumptionallowancesin non-metropolitan
areas. In the metropolitanindustrializedareasof
MexicoCity, D.F., Monterreyand Guadalajara,only 60% of the presentvalue of depreciatio
allowancescould be deductedin the first year. R&D investmenttax credit at 15% for the
purchaseof technologicalresearch(20% for smalland microenterprises),and 20% for capital
purchases by technologicalenterprises (30% for small and micro enterprises) were also
permissible.
1991-Prtsent: Effective 1991 all CEPROFI related incentives were eliminated.
However,the immediatedeductionof presentvalue of investmentexpendituresdiscountedat
7.5% per annumstill remains.
* 30 1.5.2 Tax Policy Effects on the Rental Rate and Capital
The modelwas appliedto two Mexicanindustries;detergentsand otherchemicals. The
data for thesetwo three-digitMexicanindustriesfor the period1970to 1983wascollectedfrom
a varietyof Governmentof Mexicosources. These two industriesare amongthe three largest
industriesin the industrialsector (SIC 35) comprisingof chemicals,petroleumderivatives,
rubber and plasticsproducts. Together thesetwo industriesaccountedfor 5.2 percent of total
manufacturingoutputand 2.9 percentof total employment. The data on industrycapitalstock
was developedby using the perpetualinventorymethodwith an assumeddepreciationrate of
0.08, representinga weightedaverageof assumeddepreciationrates of 0.1 for machineryand
equipmentand 0.025 for structures respectively.) Quantityof labor was measuredas the
averagenumberof employeesduring the year. The price of labor was derivedby dividingthe
total employmentcost during the year by average number of employees. Quantity of
intermediateinput, was obtainedby dividingthe cost of intermediateinputsby the input price
index.
We wil now examinethe effectsof corporatetax policyinitiativesin stimulatingcapital
expendituresin the short, intermediateand long runs for the case of Mexico. The three tax
instrumentsthat we considerfor Mexicoare the corporateincometax (CM rate, the investment
tax credit (ITC) rate, and the capital cost allowance(CCA)rate. As discussedearlier on the
theoreticaland empiricalmodels,only the relativeprice of the capitalinput is directlyaffected
by tax policy initiatives(see equations(32)-(35)). Thus, the relativeafter tax rental rate is a
crucial variable in the determinationof the effects of tax policy initiatives on capital
expenditures. In table M2 we present the elasticitiesof the tax instrumentson the rent rate.
-
31 -
Since the normalized after tax rental rate on capital is the same for both industries, the results
found for the elasticities of rental rate of capital with respect to the three instruments are also
the same. These elasticities remain relatively constant over the sample period. As seen in table
M2, a 1 percent increase in the CCA rate results in a 0.63 percent decrease in the normalized
after tax rental rate, whereas a 1 percent rise in the ITC rate leads to a 0.41 percent decline in
the relative rental rate. In fact, a 1 percent increase in the CIT rate leads to around a 1.00
percent increase in the after tax relative rental rate. The results for the short, intermediate and
long-run tax elasticities for capital demand appear in Table M3.
Table M2
Elasticities of Rental Rate of Capital with Respect to Tax Measures
Year
e
1979
-0.405
-0.621
0.895
1980
-0.409
-0.635
0.918
1981
-0.409
-0.635
0.962
1982
-0.409
-0.635
1.021
1983
-0.409
-0.635
1.021
- 32 -
Table M3
Capital Demand Elasticities
Detergents
1979
Other Chemicals
1983
1979
1983
Short-Runl
ekft
0.015
0.012
0.008
0.006
e- ,
0.024
0.019
0.013
0.009
e,,
-0.034
-0.031
-0.018
-0.014
es,
0.020
0.016
0.011
0.007
e-,
0.031
0.024
0.016
0.012
eit
4-0.045
-0.039
-0.023
-0.019
e,k,,
0.022
0.017
0.012
0.008
ekse.
0.034
0.027
0.018
0.013
-0.049
-0.043
-0.026
-0.021
Intermediate-Run
Long-Run
eu,,
-
33 -
1.5.3 Tax Incentives,InvestmentImpactsand ForegoneRevenues
Althoughfocusingon investmentexpenditureonlyprovidesa partial view of the effects
of tax policy, in this section,we calculateinvestmentinputper unitvalue of foreigngovernment
revenue. This measureis referred to as the incrementalbenefit-costratio in Table M4. These
calculationsare presentedfor the mostrecent year (1983)in the data as well as wiearlier year
(1979), together with the mean and standard deviationfor the 1979-83period. The table
suggests that the effectivenessof investment tax credit for both Mexican industries has
deterioratedin recentyears and the measureis not cost-effectivein any of the runs. Accelerated
capital consumptionallowances, have also proved to be not cost-effectivetax incentive
instrumentsas the benefit-costratio for this measureis less than one in all runs for the two
industries. Finally, while corporate tax rate reductions have had fairly large impacts on
stimulatingcapital expendituresin the detergentsand other chemicalsindustries, revenues
foregonefromsuchreductionsfar exceedthe positiveinvestmentimpactstherebyyieldinga low
benefit-costratio. Thus it is apparentthat ali three tax incentivesprovedto be cost-ineffective
in all runs for the two industriesexaminedhere.
- 34 Table M4
Investment Impacts Per Unit Value of Lost Tax Revenue
Short Run
Impact
Intermediate
Run
Long
Run
Tax Instrument
Industry
Year
InvestmentTax Credit
Detergents
Other Chemicals
1979
0.55
0.28
0.69
0.36
0.74
0.40
Detergents
Other Chemicals
1983
0.44
0.26
0.51
0.32
0.54
0.34
Detergents
Mean
(s.d.)
0.57
(0.08)
0.26
(0.02)
0.71
(0.13)
0.35
(0.02)
0.77
(0.16)
0.40
(0.03)
Detergents
Other 'hemicals
1979
0.40
0.20
0.50
0.27
0.54
0.29
Detergents
Other Chemicals
1983
0.32
0.19
0.38
0.24
0.40
0.25
Detergents
Mean
(s.d.)
0.42
(0.06)
0.19
(0.01)
0.52
(0.09)
0.26
(0.02)
0.57
(0.12)
0.29
(0.03)
Detergents
Other Chemicals
1979
0.05
0.01
0.06
0.02
0.07
0.02
Detergents
Other Chemicals
1983
0.03
0.01
0.04
0.01
0.05
0.01
Detergents
Mean
(s.d.)
0.04
(0.01)
0.01
(0.00)
U.06
(0.01)
0.01
(0.00)
0.06
(0.01)
0.02
(0.00)
Other Chemicals
Accelerated Capitol
Consumption Allowance
Other Chemicals
Corporate Income Tax
Rate Reductions
Other Chemicals
-35
-
1.6 PAISTAN
1.6.1 Tax HListory
Pakistan has followeda stablecorporatetax rate regime since the early 1960s. The
corporateincometax at 30% and a super tax at 20-25% have been maintzinedconsistently
during the last two decades. Only in the fiscal year 1989-90the super tax rte was brought
down to 15%. Foreign direct investmentreceives tax treatment equivalent to domestic
investment.Lossesare allowedto be carriedforwardsix years, but no carrybackof suchlosses
is permitted. A sales tax at 12.5%is payableon all domesticallymanufacturedgoods Dythe
producerand on importedgoodsby the importer. In the fiscal year 1989-90,import duties at
differentialrates were imposedon importedmachineryand equipment.These rates variedfrom
20% to 50% if similarmachinerywas not manufacturedin Pakistan,and a higherrate of 80%
appliedto importedmachinerywith domesticsubstitutes.
The regime of fiscal incentivesthrough the corporate income tax has experienced
significantchanges over time, as Palistan has relied upon a variety of fiscal incentivesto
stimulateinvestment. Thcse includeacceleratedcapital consumptionallowancesfor certain
physicalassets, full expensingfor R&D investments,tax rebates,regionaland industryspecific
tax holidays,and investmenttax credits. Theseare briefly discussedbelow. Furtherdetailsof
the currenttax regimeare givenin TableP1.
Tax holidays: Tax holidaysfor two years for specificindustries(e.g. engineeringgoods)
and specificregions(mostof the countryexcept majormetropolitanareas) were introducedin
1959-60. The holidayperiod was subsequentlyraised to four years in 1960-61. These tax
holidayswere eliminatedin 1972-73but reinstatedagainin 1974-75. Presentlytax holidaysfor
-36 -
five years are permittedto engineeringgoods, poultryfarmingand processing,dairy farming,
cattle or sheep breeding, fish farming,data processing,industriesmanufacturingagricultural
machinery,and also to all industriesin designatedareas of the country.
Investmenttax credits: Industriesare eligible for varying tax credits according to
location. A general tax credit for balancing,modernization,and replacementof plant and
equipmentwas introducedat a rate of 15%, but its applicationwas restrictedto designatedareas.
Since 1976-77,the creditwas madeavailableregardlessof locationand ypeof industry. This
credit was withdrawnin 1989-90but reintroducedin 1990-91.
Acceleratedcapital consumptionallowances: Capital consumptionallowancesfollow
acceleratedschedulesfor machineryand equipment,transportvehiclesand housingfor workers
(25%), oil explorationequipment(100%), ship building(20-30%),and structures(10%)on a
decliningbalancemethod. Expendituresrelating to research and development,transfer and
adaptationof technologiesand royaltiesare eligiblefor full expensing.
1.6.2 Tax PolicyEffects on the RentalRate and Capital
The r.odel was applied to the wearing apparel (SIC Code: 322) and the leather and
leatherproductsindustries(SIC Code: 323)industriesof Paldstanfor the period 1966to 1984.
The data on thesetwo manufacturingindustrieswascollectedprimarilyfrom the variousissues
of the two annual publications of the Governmentof Pakdstai.namely the Census of
ManufacturingIndustriesand the EconomicSurvey. The wearingapparel industryin 1984
contributed0.63 percentof the total manufacturingoutput and employedroughlyone percent
of the total manufacturinglabor force. The leather and leatherproductsindustry,on the other
-
37 -
Table Pl
The Structure of Corporate IncomeTax System in Pakistan 1990-91.
A. CIT ras applied to all incomecxcept dividendsand bonus shares:
30
1. Incometax rate
2. Super tax rue:
-Bakng oompmi
25
-Non-bankingcorpanies (NB)
20
3. Suchargae
10
D. CIT azeapplied to intcrcorporatcdividends(ID) & bonus shares (BS):
0
1. Income tax on D nd BS
2. Super tax on dividendsrcived by
-Domestic publio companies
5
-Poreign compania
IS
-Domeadoprivate companies
20
3. Super tax on bonus sharesissued by
-publio companiea
10
-privaescompanies
is
C. Tax rebae:
1. Tax rebateson super tax for NB public oompaniae(NBPUC)
10
2. Tax rebateson super tax for small companics
5
2
3. Tax rebata on super tax for companiesangagedin spocifc economicactivities
10-15
4. Tax rbata on income & super taxes for exportu
25-75
D. Tax Crediuson the amount of investmentin:
1. Sharel/debenturesof the Equity ParticipationFund
50
S
2. Dcbenturesnegniable bonds
3. Shars of industrialcompanie set up in Lackwardareas
10-30
4. Plant/machinery for bal., mod., repl. or extension(BMR.E)
15
E. Depreciaton AllowaLncea
1. 'Normal (annual) depreciationallowsnca (ND)
5-30
2. Extra shiAworking LUcwances(u % of ND) on plant
50-100
3. Initialdeprociationallowances
25-100
P. Pull tax holiday, raging from 4-10 yesrs, for comparies engagodin:
-manufacturinggarments
-key industrice
-anufacturing electrical equipment/itscomponents& set up in NWP
-fih' catching, cattle/sheepbreeding & dairy faJrming
-wcxpiora;5on
of specific minertl
-an indt4riLl undertakingaet up in an export processingzone
-producingdefense equipment or arnuemnt. set up in specific Lreas
-industia undertaings set up in specific backwardregions
Purta tax holidays (25-50% of the capital), ranging for 5-10 year, for oompanies set up in specific
regions and engaged in manufaucturing
goods, ship buildinp and navigadon, or generaion and supply of
electrical energy or hydraulicpower.
'
Surwhage are levied on total income and super taxes if the company's taxable income, includingdividends,
exceds Rs. 100,000.
In the cae of NBPUCs,this is an additionaltLx rebate on super tax.
Source: Ehdai, J. 1991).
-
38 -
hand, in 1984, accountedfor 1.8% of total value of output and employedone percent of
manufacturinglabor force. Together, these two industries accountedfor 2.4 percent of
manufacturingoutputin 1984.
The quantityof labor is measuredas total numberof days workedduring the year and
a labor price index was developedby dividingtotal employmentcost during the year by the
numberof days worked. The value of materialsor intermediateinputs include electricity,
petroleumfuel, naturalgas, and importedand domesticallyproducedmiscellaneousmaterials.
The quantityof materialswasconstructedby dividingthe totalvalueof materialsby an industry
level materialsprice deflator.The quantityof outputwasconstructedby dividingthe total value
of output by an industry output deflator. The series on capital stock were developedby
employingperpetualinventorymethodto investmentseriesand assuminga depreciationrate of
0.08. This representsa weightedaverageof assumeddepreciationrates of 0.1 for machinery
and equipmentand 0.025 for structuresrespectiveiy.
We now considerthe effectsof the three tax instruments;the investmenttax credit (ITC)
rate, capitalcost allowance(CCA)rate and corporateincometax (CIT) rate on the rentalrate
of capital. Table P2 showsthe emprical results we obtainedfor the elasticitiesof rental rate
of capital with respect to various tax measuresfor Palistan's wearingapparel and leather
productsindustries. The magnitudeof the ITC elasticityincreasedfrom 1977to 1984. In 1984,
a 1 percent rise in the ITC rate leads to a fall of 0.39 percentin the normaized after tax factor
price of the capital input. Over the sameperiodtime, the CCA elasticityof the relativerental
rate of capital decreased. The CIT elasticitiesdiffer slightlyacross the leatherproductsand
wearing apparel industries,but over time the elasticitiesdiffer dramatically. In the leather
- 39 -
productsindustrya 1 percentchangein the CIT rate leads to a 0.42 percentrise in 1977in the
normalizedafter tax rentalrate of capital. However,in 1984, increasesin the CIT rate result
in a rise of only 0.04 percentin the relativerentalrate. In 1977,a 1 percentincreasein the CIT
rate results in a 0.36 percent increasein the relative rental rate in the apparel industry. By
1984,a risein the CIT rate leads to a rise in the price of capitalinput of about 0.03 percentin
the same industry. The ITC elasticitiesare larger in absolutevalue than the CCA and CIT
elasticitiesin 1984althoughin 1977the CIT elasticitiesare larger than comparableelasticities
for the ITC and CCA rates. The results for the short, intermediateand long-runtax elasticities
for capitaldemandappearin Table P3.
Table P2
Elasticityof RentalRate of CapitalWith Respectto Tax Measures
Apparel
1977
-0.338
-0.285
0.359
1984
-0.386
-0.225
0.034
Leather
1977
-0.326
-0.287
0.425
1984
-0.386
-0.225
0.037
-
40 -
TableP3
Capital DemandElasticities
ApparelLehr
.Shor-Rn
e,,,,
ek,,
0.011
0.009
Xt*
-0.012
0 004
0.002
0.003
0.003
0.002
0.001
-0.004
-0.004
-0.0002
-0.021
0.008
0.005
-0.007
0.006
0.006
-0.008
0.003
0.002
-0.0003
0.046
0.029
0.016
0.006
0.038
0.017
0.014
0.004
-0.048
-0.003
-0.021
-0.0006
tadtRun
e*,,,
e.- ,0.016
e,k
LonzRun
et.,,
et.
euk
0.019
1.6.3 Tax Incentives, Investment Impacts and ForegoneRevenues
the benefitcost ratios for eachof Emetax incentivefor Pakistanare presentedin Table
P4 for the mostrecent year (1984)in the data as wellas for an earlieryear (1977),togetherwith
the meanand standarddeviationfor the 1977-84period. In carrvingout thesecalculatiors,we
note that investmentis most responsiveto changes in investmenttax credit. The loss Li
governmentrevenuesare quite similarfor ITC and CCAs,and therefore,ITC yieldsa slightly
higherbenefit-costratiothan CCAchanges.For corporatetax rate reductionslossin government
revenuesfar exceedth^ investmentimpacts.Investmentimpactsfor all measureswerc smaller
in recentyears comparedto earlieryearsfor the short and immediateruns due to the observed
decline in own price elasticity of capital in recent years. Thus the table suggeststhat the
investmenttax credit becamea cost-effectivemeasurefor boLhindlustriesin recent years based
on its long run impact only. A similar patternof cost-efftct!ivenc.ss
emergesfor accelerated
- 41 -
capital consumptionallowances. Such allowanceswere not cost-effectivein the short and
intermediaterun, and became cost effectivein recent years based on the long run impact.
Finally,corporatetax rate reductionshad very largepositiveimpactson stimulatinginvestment
on both the apparelor leatherproductsindustriesbut theseimpactswere outweighedby major
revenuelosses to the nationaltreasury. Thus for Pakistaniindustries,the three tax incentives
consideredwere ineffectivein stimulatinginvestmentin recent years but in view of a better
recordof accelerateddepreciationallowancesand investmenttax creditsin earlieryears, perhaps
a redesign of such incentiveswith some considerationfor refundabilityprovisions and
eliminationof regulatory bottleneckswou!d help restore their effectivenessin stimulating
investments.
-
42 -
Table P4
Investment Impacts Per Unit Value of Lost Tax Revenue
IstmentTax Credit
Apparel
Leather
Mea7
0.40
0.26
0.76
0.25
1.11
0.24
Apparel
Leather
198.4
0.28
0.11
0.71
0.28
2.50
2.54
Appure
Mean
(s.d.)
0.40
(0.18)
0.24
(0.22)
0.76
(0.34)
0.36
(0.32)
0.70
(2.13)
0.37
(1.44)
Apparel
Leather
1977
0.52
0.18
0.64
0.18
0.81
0.17
Apparel
Leather
1984
0.23
0.09
0.59
0.23
2.10
2.13
Apparel
Mean
(s.d.)
0.31
(0.13)
0.19
(0.18)
0.60
(0.27)
0.28
(0.26)
0.51
(1.70)l
0.25
(1. 14)
Apparel
Leather
1977
0.05
0.01
0.13
0.01
L.21
0.02
Apparel
Leather
1984
0.00
0.00
0.00
0.00
0.00
0.00
Apparel
Mman
(s.d.)
0.00
(0.06)
0.00
(0.00)
0.)4
(0.04)
0.00
(0.00)
0.08
(0.07)
0.01
(0.01)
Leather
Accelerated Capital
Consumption Allowances
Leather
Corporate Income Tax
Rate Reductions
Leather
-43
-
1.7 TURKEY
1.7.1 Tax History
Thecorporateincometax in Turkeyprovidesa significantsourceof governmentrevenues
(accountingfor 10%of total tax revenues)as well as serve as majortool of industrialpolicy.
The governmenthas changedboththe tax rate and the tax basemanytimesduring the past three
decades. The statutorycorporatetax rate hoveredaround 10%during the SO's, rose to 20%
in the 60's and to 25%in the 70's. In 1980,it was raised to 50%,loweredto 40% in 1981and
then raised again to 46% (plusa defensesurchargeof 3%) in 1985and has stayedat that level
sincethen.Overtheseyearsthere also havebeen significanttax base changes(see Bulutogluand
Thirsk, 1990).Preferentialtreatmentof publicenteprises has beeneliminatedsince 1980. Intercompany dividends distribution have been made exempt from taxation and corporate
reorganizationsare no longersubjectto capitalgains taxation. Inflationaryadjustmentof assets
but not of liabilitieshave been also allowed.
In the following,we briefly summarizethe currentprovisionsof the corporatetaxation
and investmentincentivesregimeswhich appear in Table T1. Taxable income of corporate
entities (definedas book profits before taxes plus increases in pension reserves and general
provisionfor bad debtminusinvestmentand exportallowancesand depreciationdeductionsetc.)
is currentlytaxed at a flat rate of 46%. A 3% defencesurchargeis payableon this basic rate.
In addition,a 1% tax is payableto the SocialAssistanceand SecurityFund, and an additional
1% tax is leviedfor the Apprenticeship,Vocationaland TrainingEncouragementFund, for a
combinedcorporatetax rate of 49.38%. Corporatetax is withheldat sourceat varying rates
with 0% rates for dividenddistributions,5% for incomefrom crude oil exploration,10%on
-44 -
interestand moveablepropertyincome,20% for incomefrom immoveableproperty,and 25%
for salariesand wagesand patentsand royalties.
Depreciationallowancesare based on historicalcosts adjustedby the wholesaleprice
indexminus 10% and take the form of ten-yearinterestbearingbonds. Eitherthe straight-line
or decliningbalance methodof depreciationmay be chosen for any asset, but no switch is
allowed from the straight-lineto the decliningbalancemethod during the life of the asset.
Depreciationon moveablefixedassetsacquiredon or after January1, 1983maybe takenunder
a straight-linemethodat any rate chosenby the tax payer, up to an annualmaximumof 25%.
If the decliningbalance methodis used, the maximumallowable depreciationrate is 50%.
Assetshavingvaluesless than 5,000 TL can be deducted. For structuresand moveablefixed
assetsacquiredbeforeJanuary1, 1983,the Ministryof Financepublishesmaximumdepreciation
rates (on a straight-linebasis)permissiblefor tax purposes. These rates typicallyare 4% for
factory buildings, 15% - 20% for transportequipment, and 12.5% for machineryand equipment.
A value-addedtax is leviedat a generalrate of 12%. Bankingand insurancetransactions
are subjectto a 3 % tax (BrMT).There is an investmentincentiveallowancein Turkey which
is a deductionfrom the taxableincome for corporatetax purposes. The deductionis claimed
in the year of investmenton that portion of investment which is not subsidizedby the
government. Unusedinvestmentallowancescan be carried forwardindefinitely. The rate of
investmentallowancevaries by regionand type of investment.
Corporationscan also set asideup to 25% of taxableincomefor futureinvestments.The
amountset aside at the discretionof the corporationis deductedfrom its taxableincome and
depositedin an interestbearingaccount(earningthe sameinterestas governmentbonds,usually
- 45
-
about 20% p.a.) with the CentralBank. It can be withdrawnany time with authorizationfrom
the State PlanningOfficeand used for investment.
For tax purposes, capital is depreciatedat a rate of up to 50% for machineryand
equipment. Further assetscan be revaluedat the end of every calendaryear.
A largenumberof non-taxincentivesare availableto eligibleinvestments.Theseinclude
low interestcredit, fundsfor workingcapital,allocationof foreignexchange,and allowancefor
importof used equipment.
-46
-
Table Ti
The Structure of Corporate Income Tax System in Turkey 1990/91
(Figures in percent)
Corporate Income Tax: General
Withholding tax rates on payments by a domestic
corporation to a foreign corporation
Rental from fixed assets
Leasing
Royalties on patents
Professional services
Petroleum services
Interest on trade receivables
Other interest (loans and deposits)
Withholding taxes on payments to nonresident individuals
Rentals from immovable assets
Royalties on patents
Services (professional)
interest on receivables & deposits
Value-added tax
Standard rate
Agricultural product
Basic foods, books, natural gas
Luxury goods
Petroleum products
Banlkng and Insurance transactions tax
Investment incentive allowance
Export allowance
Export earnings of manufacturer
Export earnings of traders
Export of fresh truit, vegetables
International Transport
Tourist establishments
46
20
0.5
25
15
S
10
10
20
25
15
10
12
1
6
20
13
5
30-100 of the cost of
specified assets
12
3
12
12
20
DepreciationAlowance
Straight-line
Declining-balance
Source: Price Waterhouse (1992)
25
50
-
47 -
1.7.2 Tax PollcyEffects on the Renl Rate and Capital
The model Is appliedto three Turldshindustries;non-electricalmachinery(SIC 382),
electricalmachinery(SIC383)and transportequipment(SIC384)industriesin the privatesector
only and covers the period 1973 to 1985. These industries accountedfor 20% of total
manufacturingoutputand employmentand 24% of manufacturingwagesin 1985.The data on
output, employment,intermediateinput and investmentwere obtained from a variety of
Governmentof Turkeysources. The quantityof labor was measuredas the averagenumberof
employeesduring the year. The price index was constructedby dividingtotal employmentcost
duringthe year by averagenumberof employees. Intermediateinputsor materialsincluderaw
materials, components, containers, fuel and electricity. The quantity of materials was
constructedby dividingtotal valueof materialsby an industrymaterialsdeflator. The quantity
of outputwas constructedby dividingthe total valueof outputby the relevantindustryoutput
price deflator. The same deflator was used both for the electrical machineryand transport
equipmentindustries. The capitalstockseries weredevelopedby applyingperpetualinventory
methodto investmentseries and by assumingdepreciationrate equal to 0.08, representinga
weightedaverageof assumeddepreciationrates of 0.1 for machineryand equipmen and 0.025
for structures.1
The effectsof the three tax instrumentson the rental rate of capital are given in Table
T2. Sincethe normalizedafter tax rentalrate on capitalis the samefor the three industries,the
results foundfor the tax elasticitiesare also the same. From tableT2, we observethat the IA
elasticity increases over the sampleperiod, whereas the CCA and CIT elasticitiesremain
relativelyconstant. Over the first half of the sampleperiod, a 1 percentincreasein the IIA rate
-
48 -
decreases the afkr tax rental rate by 0.20 percent. Over the second half of the period, the
elasticity ranges from -0.24 to -0.35. For most of the period the elasticity associated with the
CIT rate ranges from 0.21 to 0.28 and then decreases of the last few years. Generally, the
CCA rate clasticity ranges from 0.70 to 0.10 for most of the period. The results for the short,
intermediate and long-nm tax elasticities for capital demand appear in Table T3.
Table T2
Elastcities of Rental Rate of Capital With Respect To Tax Measures
._
_
_
_
1973
-0.199
-0.065
0.210
1974
-0.195
-0.086
0.242
1975
-0.196
-0.084
0.238
1976
-0.199
-0.067
0.212
1977
-0.197
-0.078
0.229
1978
-0.193
-0.098
0.260
1979
-0.193
-0.096
0.259
1980
-0.242
-0.129
0.386
1981
-0.348
-0.147
0.259
1982
-0.345
-0.155
C.276
1983
-0.258
-0.064
0.057
1984
-0.258
-0.063
0.055
1985
-0.341
-0.099
0.101
-
49 -
Table T3
CapitalDemand Elasticities
1974
1985
1974
f1988
ewk,,
0.014
0.013
0.024
0.021
0.024
0.020
e-,,
0.006
0.004
0.010
0.006
0.010
0.006
eb,,
-0.017
-0.004
-0.029
-0.006
-0.029
-0.006
0.021
0.037
0.033
0.037
0.032
0.009
0.006
0.016
0.009
0.016
0.009
-0.027
-0.006
-0.046
-0.009
-0.046
-0.009
eti,
0.034
0.034
0.059
0.052
0.055
0.051
e-,
0.015
0.009
0.026
0.015
0.026
0.015
Ckck
0.042
-0.009
-0.074
-0.015
-0.074
-0.015
.....
~~~i
1974 i
74
19
...............
.........
.
ShorRn
Intomediate
&a
ek. ak^;0.021
Je-.
3et.,,
Long-Run
1.7.3 Tax Incentives, Investment Impacts and Foregone Revenues
Table T4 presents the benefit-cost ratios for the three Turkish industries for two years
1975 and 1985 and the mean and standard derivation for the sample period 1975-1985. A 1%
Lhicase in investment allowance (IIA) had the largest effect on capital while a similar change
in capital consumption allowances (CCAs) and corporate tax rate reductions had relatively
smaller impacts. This is because the elasticityof rental rate of capital with respect to investment
allowances is much higher than with respect to capital consumption allowances and corporate
tax rate reductions. The loss in tax revenue associated with tax rate reduction are quite large
-
50 -
and therby yieldinga low benefit-costratio for sucha policychange. The revenuelosses are
largerfor theinvestmentallowancethanfor changesin capitalconsumptionallowancesand since
investmentimpactsare higherfor the formermeasure,the net effectis to yield similarbenefitcost ratios for the two measures. The benefit-costratio is smaller for almost all measuresin
1985relativeto 1975. This results from a declinein the elasticityof capitalstock to a change
in its own rental rate. Note that the capitalstockincreasesover time, implyingthat if the own
price elasticityof capital were to be constant,investmentresponseto changesin rental rate
would have to increaseat the same rate as the increasesin capital stock. It is unlikelythat
investmentresponsewill increaseat the samerate becauseit wouldimplyan unrealisticincrease
in the marginalproductof capital. Thus it is reasonableto expectown priceelasticityof capital
to decline over time. In conclusion, the table suggests that investment allowancesand
accelerateddepreciationprovisions proved to be effectiveinstrumentsof public policy for
investmentpromotion,especiallybased on their intermediateand long run impacts. The same
couldnot however,be saidaboutcorporatetax rate reductionswhichclearlyresultedin windfall
gains to existingcapital withoutencouragingnew investment.
Table T4
Invoitmen hmpctdsPer Unk Value of Lost Tax Revenue
: ;.
; jj:
nvestinet Allowance
i i jjji;*
. .;
R .
Electrical Machinery
Non-Elect-cl Mchinery
Trnapon Equipment
1975
0.63
1.00
1.14
0.97
1.59
1.71
1.50
2.62
2.56
Electial Machinery
Non-EloctricalMachinery
Tranport Equipmet
19U.
0.40
0.86
1.00
0.72
1.42
1.54
1.54
2.49
2.40
Electrical Machinery
Mean
(s.d.)
0.53
(0.012)
0.81
(0.17)
0.85
(0.23)
0.84
(0.17)
1.29
(0.28)
1.34
(0.35)
1.37
(0.29)
2.12
(0.51)
2.19
(0.60)
ElBouia Machinery
Non-Electio Machinery
Transport Equipmet
1975
0.56
0.89
1.01
0.86
1.42
1.53
1.33
2.34
2.28
Elebia Machinery
Non-Electic Machinery
TanAport Equipment
1985
0.38
0.51
0.94
0.6"
1.33
1.44
1.45
2.34
2.25
Electrical Machinery
Mean
(s.d.)
0.47
(0.10)
0.72
(0.14)
0.76
(0.20)
0.75
(0.14)
1.15
(0.23)
1.20
(0.31)
1.22
(0.24)
1.39
(0.43)
1.94
(0.51)
erical Machincry
Non-Blootl Machinay
Tranport Equipment
1975
0.32
0.16
0.20
0.56
0.27
0.31
0.84
0.45
0.50
EletoariclMachinery
Non-ElectricalMachinery
Traupot Equipment
1985
0.20
0.07
0.03
0.21
0.11
0.06
0.28
0.19
0.10
ElectricalMachinery
Mean
(s.d.)
0.06
(0.36)
0.05
(0.37)
0.08
(0.71)
0.01
(0 28)
0.03
(0.51)
0.02
(0.23)
0.00
(0.45)
0.07
(0.38)
0.12
(0.96)
Non-Electrica Machinery
Transport Equipment
c apiwal
naumptionAllowUanc
Non-Electical Machinery
Trasponation Equipment
orporale IncomeTax
Reduction
Non-ElectricalMachinery
Transport Equipmenz
52 1.8 SUMMARYAND CONCLUSIONS
This paper providesan empiricalframeworkfor the assessmentof tax policyeffectson
the array of producer decisionsconcerningoutput suppliesand input demandsin Mexico,
Paldstanand Turkey. A dynamicproductionstructuremodelis specifiedand estimatedfor this
purposefor selectedindustriesin eachof the count'. s.
On the Elasticityof RentalRateof Capitalwith Respt to Tax Instruments:The tax sensitivity
of rental rate of capitalis quite inelasticwith the singleexceptionof its elasticitywith respect
to corporatetax rate in Mexicowhich is unitary(see Table SI). In Mexico,the rental rate of
capitalis most sensitiveto corporatetax changesand relativelyless to accelerateddepreciations
and investment tax credits. In Pakistan, the sensitivityranking of three instruments is
completelyreversedand investmenttax creditchangeshave the greatestinfluenceon the rental
rate of capital. In Tlrkey, the rental rate is more responsiveto changes in investment
allowancesthan acceleratedcapital consumptionallowances(CCAs)or the corporatetax rate
reductions.
TableSI
Elasticityof Rental Rateof Capitalwith Respectto Tax Measures
Mexico(1983)
-0.409
-
-0.635
Pakistan(1984)
-0.386
-
-0.225
1.021
0.035
Turkey (1985)
-
-0.099
0.101
-0.341
-
53 -
On the Tax Sensitivityof the CapitalStock: The capitalstock exhibitssensitivityto tax changes
but this sensitivityvaries by tax measure,by industryand by the adjustmentperiod. TableS2
providescomparativeevidenceon the tax sensitivityof the capital stock by industry, by tax
measure, and by adjustmentperiod. For Mexico, elasticityestimatesrange from -0.014 to
-0.043 for corporatetax changes;from 0.009 to 0.027 for CCAs;and from 0.006 to 0.017 for
changesin investmenttax credits. For Palistani industries,the responsivenessof capitalstock
to changesin corporateincometax is quite small - elasticityestimatesrange from 0.0002 to
-0.006; for investmenttax credit elasticityestimatesrange from 0.002 to 0.029; and finallyfor
capitalcostallowancesbetween0.001 and 0.017. Thelast twosets of elasticitiesare compatable
withthe onesobtainedfor theMexicanindustries.For Turkishindustries,changesin investment
allowancesmatter more for the effects on capital formation than alternate tax measures.
Specifically,elasticityestimatesrange from 0.013 to 0.052 with respect to changes in the
investmentincentiveallowance;from 0.004 and0.015 withrespectto changesin the capitalcost
allowances;and from -0.004 to -0.015with respectto changesin the corporateincometax.
On Benefit-CostRatios: The model results suggestthat tax policy affected productionand
irkiestmentand further that some tax incentiveswere more effectivethan others in investment
stimulationper governmentrevenue loss (see Table S3). Among the incendves measures
examined, investmentallowancesproved to be a cost-effectiveinstrumentfor investment
promotiononly to Turkish industries;and investmenttax credits and accelerateddepreciation
provisionshad a mixed success while corporate tax reductions met with dismal failure in
promotinginvestmentin a cost-effectivemannerin all cases for all countries. In terms of their
-54 T"
S2
TaI Samiiviltyof Capial Stock
_
S
-,
~~,
AW. _K
~~.
-~~~~~i-~
MEXICO(1983)
Dctcrgeia
.012
.016
.017
-
Otherchemaical
.006
.007
.008
AppaDe
.004
.001
LAherd i
.002
ElecricalMach.
-
Noa-electrical
Mach.
Tansport Equipment
.019
.024
.027
.031
.039
.043
-
.009
.012
.013
-.014
-.019
-.021
.029
-
.002
.005
.017
-.0004
-.000
-.003
.003
.006
-
-
-
.001
.002
.004
-.000
-.000
-.006
-
-
.013
.04
.034
.004
.006
.009
-.004
-.006
-.009
-
.021
.033
.032
.006
.009
.015
-.006
-.009
-.015
-
.020
.032 .051
.006
.009
.015
-.006
-.009
-.015
-
PAKISTAN(1914)
TURKEY (1915)
-
55 -
Table S3
InvestmentExpendituresper Unit Value of Lost Tax Revenue
Short Run
Tax Instrument
InvestmentTax Credit
Mexico:DetergentsIndustries
Mexico:Other ChemicalIndustries
Pakistan:ApparelIndustries
Pakistan:LeatherIndustries
Imoact
Intermediate
Run
0.44
0.26
0.28
0.11
0.51
0.32
0.71
0.28
0.S4
0.34
2.50
2.54
AcceleratedCapitalConsumptionAllowances
Mexico:DetergentsIndustries
Mexico:Other ChemicalsIndustries
Pakistan:ApparelIndustries
Pakistan:LeatherIndustries
Turkey: ElectricalMachineryindustries
Turkey: Non-ElectricalMachineryIndustries
Turkey:TransportIndustries
0.32
0.19
0.23
0.09
0.38
0.81
0.94
0.38
0.24
0.59
0.23
0.68
1.33
1.44
0.40
0.25
2.10
2.13
1.45
2.34
2.25
CorporateIncomeTax RateReductions
Mexico:DetergentsIndustries
Mexico:OtherChemicalsIndustries
Paidstan:ApparelIndustries
Paldstan:LeatherIndustries
Turkey: ElectricalMachineryIndustries
Turkey:Non-Elctrical MachineryIndustries
Turkey: Transport.ndustries
0.03
0.01
0.001
0.00
0.20
0.07
0.03
0.04
0.01
0.0002
0.00
0.21
0.11
0.06
0.05
0.01
0.007
0.00
0.28
0.19
0.10
InvestmeneAllowance
Turkey: ElectricalMachineryIndustries
Turkey:Non-ElectricalMachineryIndustries
Turkey:TransportEquipmentIndustries
0.40
0.86
1.00
0.72
1.42
1.54
1.54
2.49
2.40
Long
RunRun
-56 -
long-runimpacts, investmnent
tax creditswere cost-effectivein two of the four industriesstudied.
Accelerated capital consumption allowances also registered a similar performance and had
incremental benefit-cost ratio exceedingone in the long run for five out of seven industries
studied. Corporate tax rate reductions stimulated investmentsbut resulted in revenue losses
exceedingthis stimulativeimpact in all cases and in all runs consideredin this study. Note that
corporate tax rate reductions apply to a larger base of pre-tax profits than the smaller base of
current investmentsrelevantfor investmnent
tax credits. The long run cost-effectivenessof these
incentives,except corporatetax rate reductionswhichproved cost-ineffectivein all cases, vares
by country. In Turkey, investmentallowancesand capital consumptionallowanceswere costeffective. In Mexico, both investrnenttax credit and acceleratedcapitalconsumptionallowances
were not cost-effective.In contrast, in Pakistan, both the investnent tax credit and accelerated
capitalconsumptionallowanceswere cost-effective.In the intermediaterun, definedas tax policy
impact after one year, only the investmnentallowances and accelerated capital consumption
allowancesavailable to Turkish industriesproved cost-effective.
In conclusion, selective tax incentives such as investment tax credits, investment
allowancesand acceleratedcapitalconsumptionallowancesare morecost-effectivein promoting
investment than more general tax incentivessuch as corporate tax rate reductions. In order to
make selectivetax incentivesmore effective,investmenttax creditsmust be refundableand carry
forward of investment and depreciationallowancesbe permitted. If stimulationof investment
expenditureis the sole objectiveof tax policy, corporatetax rate reductionis not a cost-effective
instrumentto achieve this objective.
-
57-
Enldootes
1.
The model can be readily generalizedto include multipleoutputs. The production
functionis also assumedto be twicecontinuouslydifferentiable,and quasi-concavein the
inputsand net investments.
2.
The issue of capital utilizationis not addressedin this model. The problemof costly
capitalutilizationimpliesthatdepreciationrates dependon prices, technologyand market
structure. Hencethe use of existingmeasuresof capital stockswouldbe inappropriate
becauseservicelives are assumedto be independentof prices and technology. Costly
capitalutilizationimpliesthat capitalstock measurementand technologydetermination
must be modelled simultaneously. This is an interesting, complex, but secondary
problemto determiningthe effects of tax policy on outputsupplyand input demand.
3.
We abstractfromintroducinginvestmenttax allowances.The modelcan be modifiedto
includethem alongwith capitalcost allowances.
4.
This assumptionis the Mortigliani-Miller
hypothesis.It is also possiblethat with market
imperfectionsfirmscan influencethe rate of returnon theirfinancialcapital(see Steigum
[1983]and Bernsteinand Nadiri [1986]for dynamicmodelsin this context).
5.
The formulafor the after tax purchaseprices of the capital stockscan be simplified.If
the discountrates are not expectedto changethen
Q,j
qgs(1 -V,,-(i
t-0
u+,.(l -T 1,v1 ,)d 1 ,)/(1
+
p)'). If, inaddition, thetaxrates
and
credits are not expectedto changethen Q - q4(l
-
v, - u,(l - 40v,( S d, / (1 +
-t-O
The latter is the more standardformulaand is a specialcase of the after tax purchase
price formulaused in the model (see Hall and Jorgenson[1967, 1969]and Arrow and
Kurz (1970)).
6.
The inverseprice elasticityand the conjecturalelasticityare not assumedto be constant.
Equation(12.1) containstheir equilibriummagnitudes.The productionfunctionis also
part of the first order conditions.The second order conditionsare assumed to be
satisfied.The symbolV representsthe gradientvector.
7.
Recall that (<Oaide>o so the last set of terms on the right side of equation (14),
includingthe minussign is positive.
8.
The additionalrevenueand therebyprofit arising from oligopolypower does not vary
when it is evaluated at the equilibrium point. Thus the term affects the calculation of
variableprofitbut doesnot affectthe first order conditionscharacterizingan equilibrium
As a consequencethe expressioncan be ignoredwhen definingshadowvariableprof'
-
58-
9.
The functionis also twice continuouslydifferentiable,homogeneousof degreeone and
convex in after tax prices, and concavein the capitalinputs and net inventmentlevels.
10.
It is also assumedthat the tansversalityconditionsare satisfied.The symbol0. signifies
a m dimensionalvector of zeros.
11.
Sincedepreciationrates for the sampleindustriesare not available,Jorgensonand Yun
(1991)estimatesfor U.S. industrieswere used. The depreciationrate for non-residential
structures(0.025)was calculatedas an averageof the depreciationrates on varioustypes
of industrialstructures. Inclusionof othertypes of buildingsand structuresdid not alter
rate significantly. The depreciationrate for producerdurable
the above epron
equipment(0.10) was calculatedas an averageof the depreciationrates on a numberof
electrical,non-electricaland transportationmachineryand equipmentcategories.
Pleasenote that whileVariousstudiesuse a rangeof depreciationrates, theseare similar
to the rates assumedhere. For example,Epsteinand Yatchew(1985)use a figure of
0.107 and Epsteinand Denny(1983)use a range between0.102 and 0. 111.
12.
See endnoteNo. 11.
13.
See endnoteNo. 11.
-
59 -
References
Arrow, K., and M. Kurz. 1970. Publc investment,the Rste of Return and ODtimalSocial
Pogicy,Baltimore,JohnsHopkdnsPres.
Bernstein,J.I., and M.I.Nadiri. 1986. "Financingand investmentin plant and equipmentw,in
M.H.Pestonand R.E.Quandtad., Pries. Co petitin o,
OxfordEngland:
Philip Allan/Barnes& Noble.
Bernstein,J.I., and M.I. Nadiri. 1988. "Corporatetaxes and the structureof production,a
selectedsurvey, in J. Mintz and D. Purvis ed., The Impa of Taxationon Business
AcdI3i, KingstonCanada:John DeutschInstitute,Queen's University.
Bernstein,J.I., and P. Mohnen. 1991. "Pricecost margins,exportsand productivitygrowth,
with an applicationto Canadianindustries', CanadianJoural of Economics.,24, 638659.
Boadway,R. and A. Shah. 1992. The Roleof InvestmentIncentivesin DevelopingCountries.
Forthcomingin Anwar Shah, editor, Fiscal Incentivesfor Investmentin Develoin
Conties, WorldBank, Washington,D.C.
Bulutoglu,Kenan, andf WayneThirsk. 1991. 'Tax Reformin Turkey'. Public Economics
Divisio,WorldBank, Washington,D.C. Processed.
Diaz, Gil-Diaz. 1990. Reformingtaxes in DevJ1opingCountries:Mexico's ProtractedTax
Reform. Processed,CountryEconomicsDepartment,WorldBank
Ehdaie, J. 1991. RevenueGeneratingand InvestmentAspectsof the CorporateIncomeTax
Systemin Pakistan:An Agendafor Raform.Processed.CountryEconomicsDepartment,
WorldBank.
Epstein, L.G. 1981. 'Duality theory and functionalforms for dynamic factor demands%,
Rejew of EconomicStudies,48, 81-95.
Epstein, L.G. and M.G.S.Denny.1983. 'The MultivariateFlexibleAcceleratorModel: Its
EmpiricalRestrictionsand an Applicationto U.S. Manufacturing.' Economea, v.52,
pp. 647-674.
Epstein,L.G. and A. J. Yatchew. 1985. 'The EmpiricalDeterminationof Technologyand
Expectations: A SimplifiedProcedure,' Ioumal of Econometrics,v.27, pp. 235-258.
Guisinger, Stephen and Associates. 1985. Investment incentives and Performance
Ruiiments. PraegerPublishers,New York.
-60Hall, R.E., and D.W. Jorgenson. 1967. 'Tax policy and investmentbehavior", American
E-conomicReview.57, 391-414.
Hall, R.E., and D.W. Jorgenson. 1969. "Tax policy and investmentbehavior, reply and
further results", AmericanEconomicReview,59, 388-401.
InternationalBureau of Fiscal Documentation:1988. Tax SystemProfiles. Mexico,Pakdstan,
Turkey. VariousIssues.
Jorgenson,D. and K. Y. Yun. 1991. Tax Reformand the Costof Capital. OxfordUniversity
Press, 1991.
MexicoD.F., Bancode Mexicovariousyears. Sistemade CuentasNacionalesde Mexico.
Mexico,Bancode. 1986. 'La Encuestade Acervos,Depreciaciony Formacionde Capitaldel
Banco de Mexico." Documento Interno, Reporte Metodologico, Direccion de
InvestigacionEconomica.
Mexico,Bancode. 1986The MexicanEconomy
Mexico,D.F. Bancode. 1987. IndicadoresEconomicos.
Mexico, Bancode. 1988. 'Enquesta Semestralde Coyuntura,2e semestre1987." Direccion
de InvestigacionEconomicaGerenciadel SectorReal. Febrero de 1988.
Mexico,Governmentof. Ministryof Finance. UnpublishedTax Data.
Mortensen,D. 1973. 'Generalizedcosts of adjustmentand dynamicfactordemandtheory",
Econometriga,_41,
657-666.
Pakistan, Governmentof. Federal Bureau of Statistics.Census of ManufacturingIndustries.
VariousIssues. Karachi
Pakistan,Governmentof. FinanceDivision. Economic
Islamabad
StatisticalSupplement:1987-88.
Pakistan,Governmentof. FinanceDivision. PublicFinanceStatistics. VariousIssues
Price Waterhouse. 1992. CorporateTaxes: A WorldwideSummary. New York.
Steigum,E. 1983. 'A financialtheoryof investmentbehavior",Eonometnca. 53, 637-6451
Treadway, A. 1971. "On the rationalmultivariateflexibleaccelerator",Econometrica
.39,
845-1355.
-
61 -
Tradway, A. 1974. "The globallyoptimalflexibleaccelerator', Iournalof EconomicTheory,
7, 17-39.
TUSIAD1987. "a TuridshEcnomy.
Turkey, Governmentof. State Institute of Statistics. Statistical Yearbook of Turkey.
PublicationNo. 1250. Variousyears.
Turkey,Governmentof. TreasuryDcVartment.UnpublishedTax Data.
Ugarte, Fenado Sanchez. 1988. 'Taxation of Foreign Investment in Mexico: The North
Ameican Perspective.' InternationalEconomicsProgram Worldng Paper DP88-7.
Universityof Toronto. Processed.
Policy Research
Working Paper Series
Contact
for paper
Author
Date
WPSI 178 Productivity
of PublicSpending,
SectoralAllocation
Choices,and
Economic
Growth
JohnBaffes
AnwarShah
September1993
C. Jones
37699
WPS1179 Howthe MarketTransitionAffected
ExportPerformance
in the Central
EuropeanEconomies
BartlomiejKaminski
September1993
P. Kokila
33716
andTaxationof U.S.
WPS1180The Financing
Direct InvestmentAbroad
HarryHuizinga
September1993
R. Vo
31047
WPS1181 ReformingHealthCare:A Casefor
Stay-WellHealth Insurance
Zeijko Bogetic
DennisHeffly
September1993
F. Smith
36072
WPSI 182 CorporateGovernancein Central
and Eastern Europe:Lessonsfrom
AdvancedMarket Economies
CherylW. Gray
RebeccaJ. Hanson
September1993
M. Berg
31450
WPS1183 Who WouldVote for Inflationin
Brazil?An IntegratedFrarnework
Approachto Inflationand Income
Distribution
CheikhKane
Jacques Morisett
September1993
T. Hollestelle
30968
WPS1184 ProvidingSocial Benefitsin Russia: SimonCommander
RichardJackman
Redefiningthe Rolesof Firmsand
and Government
September1993
0. del Cid
35195
WPS1185 ReformingHungarianAgricultural
Trade Policy:A Quantitative
Evaluation
MorrisE. Morkre
DavidG. Tarr
September1993
N. Artis
37947
WPSI186 RecentEstimatesof CapitalFlight
Stijn Claessens
David Naude
September1993
R. Vo
31047
WPSI 187 How ShouldSovereignDebtors
RestructureTheir Debts?Fixed
Interest Rates, Flexible Interest
Rates,or Inflation-indexed
AndrewWarner
September1993
J. Queen
33740
WPSI 188 Deveiopmentalism,
Socialism,and
MarioMarcel
Free MarketReform:Three Decades AndresSolimano
of IncomeDistributionin Chile
September1993
S. Florez
39075
WPS1189 Can CommunistEconomies
TransformIncrementally?China's
Experience
September1993
PRDTM
37471
September1993
T. Ishibe
37S65
Title
Alan Gelb
Gary Jefferson
InderjitSingh
WPS1190 The Government'sRole in Japanese YoonJe Cho
and KoreanCredit Markets:A New
ThomasHellmann
InstitutionalEconomicsPerspective
Policy Research Working Paper Series
Contact
for paper
Author
Data
GeoffreyJ. Bannister
September1993
A. Daruwala
33713
WPS1192 Effectsof Tax Reformon Argentina's Jacques Morisset
Alejandrolzquierdo
Revenues
September1993
G. Carter
30603
WPS1193 The ArmenianLaborMarketin
Transition: Issuesand Options
MilanVodopivec
WayneVroman
September1993
S. Florez
39075
WPS1194 How Fast HasChineseIndustry
Grown?
TomRawski
September1993
E. Khine
37471
WPS1195 The EnterpriseSectorand
Emergenceof the PolishFiscal
Crisis, 1990-91
Mark Schaffer
September1993
E. Khine
37471
WPS1196 CorporateTax Structureand
Production
Jeffrey Bernstein
AnwarShah
September1993
C. Jones
37699
WPS1197 Determinantsof Inflationamong
FrancZoneCountriesin Africa
Bruno Boccara
ShantayananDevarajan
September1993
C. Jones
37699
Title
WPS1191 Rent-Sharingin the Mufti-Fibre
Arrangement:The Caseof Mexico