Policy, Researci, and External Atiairs
WORKING PAPERS
Macroeconomi' Adjustment
and Growth
Country Economics Department
The World Bank
March 1991
WPS 631                         (
The Macroeconomics
of the Public Sector
Deficit
The Case of Morocco
Riccardo Faini
Growth has remained relatively high in Morocco, and inflation
subdued. Morocco has made great progress toward macroeco-
nomic and fiscal stability, but the need remains for an unshaken
commitment to fiscal discipline, a determined effort to reformn
the tax and public spending systems, and a measured attempt to
make credit available for investment and to liberalize financial
markets.
rhc lolic! . Rcse4rch. and I.xitcnI AAffairs ('ompk,x d,'in ., 'ls I' I Wsrh' In 3p.r s    mn'aC I5C I  Jd:ags O, Es*rk nprgrcss and
In cncouragc iho c,h. idc  d:l1': F  I I:d I  ,:.!   d    ..:  parv'g''  (a n I, the names of
tC auIthnr'. refl . t.,t kthr s:1 ,  . a': .!   Ii..! h...d  .t! a.!  % .  .. ! d :   I'  :  .::  - . a,td I. n st;oir  arc thc
o-h rrs na ,' I '    :'  d,,,d    I-t hv  :.............. ..:  ' H : U. '¾i  s. I )1r. :.'  :("'aa','v-  w a,'s  : g0  rc:cnhct couninc



Polizy, Research, and External Affairs
;I IU3s _-
Macroeconomic Adjustment
and Growti
WPS 631
'I'his papcr --- a product of' Itlc Macroeconomic Adjustment and (irowth Dlivision, Countr\ ILconomn ics
Department -- is parn ol'a PRK research project, 'l'hc Macroeconomics ol lhc Public Sector- Delicit (RI")
675-31).  Copies are available f'ree flrom tlhe World Biank, 1818 li Strect NW, Washington 1)DC  (0113.
Please contact Raquel l.u/, rootm N 1 1-057, cxtenisionl 34303 (39 pa-Cs).
Morocco's reccnt cconomic history rcsemlblcs              Faini argues that  ;Iage moderation an1d
those of'mansv At'rican countries. Morocco's          judicious monetary policies were instrumnental in
economic dilf'iculties originated in the comiminod-   restrainiig inflation. Wilih a brief exception in
ity (phosphate) boom of'tlic mid-1970s, which          1983, monetary autlhoNtics remained firmilv
coincidcd with 1-isinig govenmeint spending and       committed to avoiding inflationary financing of
an unprecedented expatnsion of'public invest-         the budget delicit, But this sitrategy could
ment -- cndilg Morocco's earlier liscal conlser-      succeed only becausc of the A ide-ranging system
vatisilm. A sudden revcrsal of the temis ol' trade    of credit and monetary rcoulationis thalt chiain-
in tie late 1970s -- a resull of'a plunge in         ncied domestic f'unids toward thc trcasIry' al
phosphate prices and thc second oil shock-            relatively low cost. But the prospects for conl-
prompted Mor. 'o to resort increasingly to            tinuing such a stratcey are not f'avorable.
external capital markets to maintain an unabated
level of public spending.                                 Grolth perforlimaince canl be attributed to an
outstanidinlg export responise to thc new trade
But the continued deterioration ol' thc terms     regimc anl(d to F1avorable supply shocks --
of trade and the unexpected rise in international     includinig a strinlg ol' record agricultural harvests
interest rates, together with thc severe droughlt of'    and thc collapse of' real oil prices.
1980-84, croded debt service capacity and
precipitated a major forcign exchange crisis in           Morocco has made gr;eat pro-gress towvard
1983.                                                 macroeconomic and liscal stability but the autlio
recommends:
In response to this crisis, Morocco launchied
a medium-term program of economic rceform  and           * An unshaken commimntcii to Fiscal disci-
introduced comprehelisive stabilization and           pline. Increa,sed govermcint spending will
structural adjustment measures. Since 1983,           probably crowd out investment, The shor-m--un
Morocco has made great progress in alleviating        benclets on output of suchl spendiiing may be
botn initernal and extcrnal disequilibria -           outweighed by its long-runi negative impact on
reducing the budget deficit from 9 percent of'        growth.
GDP in 1982 to 4.5 percent in 1988, and the              * A deteimined el'lfol to refborm thc tax and
current account deficit from 12 percent of GDP        public expcnditure system, so ihe brunit of liscal
to 0.4 percent in the same period.                    adjustment wnill not again fall mostl) on public
investment.
Initerestingly, growth has remained Flairly          * Encouraging thc availability ol' credit, which
high in Morocco, at least in relation to other        significantily inilueinces the demand for invest-
highly indebted countries, and inflation subdued.     ment.
Morocco's perfonrancc seems to contradict the            * Studying thie imnpact ol' macrocconloilic
perceived wisdom  that large budget deficits will     equilibria, especially oi tihe government budget,
foster inflation. The inflation record is particu-   to assess the best spee(I lfor finalncial liherali;a-
larly surprising because Morocco achieved a 20        tion.
percent real depreciation in the 1980s.
Thc l'RE Working Ptaper Series dissvninates the findings or \kork under w\av in th. Bank's Po h.\, kewirk h and E\toniia
AffairsCompte\. An oh jictivc ofthl s'rics is lo ge tth.sac fmnhngs out              thuicktv. e en i rpre. eorains a;. t. rtan ftlltv pot lie d
The rindfinvs. intermretins. ncf:md conchiviions in rhese p.IVT'rs d(o not neee-(:riTV represent olh ial 11:ink p,th>
f1r,.1 ut,\,,   I\s  tie  I'Rft. 'lli !lI':. ('t',:'



The Macroeconomics of the Public Sector Deficit
The Case of Morocco
by
Riccardo Faini*
Table of Contents
1.  introduction                                                           1
2.  The Dudget Deficit: Evolution and Financing                           3
a. The overall trend                                                  3
b. The primary deficit                                                5
c. Financing the budget deficit                                       7
3.  Assets Demand, Seigniorage, and the Inflation Tax                     10
4.  Investment and Saving Decisions                                       17
a. The investment choice                                              17
b. The saving decision                                               21
5   Modelling the Impact of Fiscal Policy                                25
a. A macroeconomic model                                             25
b. Simulating the impact of fiscal policy                            28
6.  Conclusions                                                          36
Endnotes                                                                 37
References                                                               38
*     This paper is an offspring of joint work with John Porter and Sweder van
Wijnbergen. I am grateful to both of them for stimulating discussions and suggestions. I
would like to thank Klaus Schmidt-Hebbel and Tobias Muller for providing me with
crucial data and a preview of their paper which has been a constant source of ideas and
insights. I am finally grateful to Roberto Fumagalli for very skillful research assistance.



1
1. INTRODUCTION
Morocco's recent economic hlstory bears strlking resemblances with
the one oF many countries in Africa. The origins of Morocco's economic
difficulties can be traced to the commodity (phosphate) boom of the mild-
1970's which coincided wlth rislng government expenditure and an
unprecedented expanslon of the public investment program signallng the
end of the conservative Fiscal policies of the past. The sudden reversal
in the terms of trade in the late 1970's as a result of the plunge In
phosphate prices and the second oil shock prompted Morocco to resort
Increasingly to external capital markets In order to malntaln an unabated
level of public expenditure. However the continued deterloratlon of the
terms of trade, the unanticipated rise in International Interest rates
together with the severe drought In 1980-84 eroded debt service
capacity and precipitated a major foreign exchange crlsls In 1983. In
response to this crisis, Morocco launched a medium-term program of
economic reforms and, In consultation with the International Monetary
Fund and the World Bank, introduced a comprehensive set oF stabllizatlon
and structural adjustment measures. Since 1983 Morocco has made
great progress in allevlating both Internal and external disequllibria.
The overall budget and current account deficits have been reduced from
9% and 12% of GDP in 1982 tc 4.5% and .4% respectlvely In 1988.
There are several interesting features In Morocco's adjustment
experience. Despite the severity of the crisls and the size of the
adjustment undertaken, growth has remained fairly high, at least In
relation to other highly indebted countries, and inflatlon subdued. As
measured by the CPI, inflation was equal to 6.2% in 1983, It Increased
to 12.5% in 1984 but then steadily declined to reach 2.3% in 1988.
During the same period real GDP growth averaged 4.3%. This
performance seems to contradict the received wisdom that large budget
deficits will foster inflation. The inFlation record Is particularly
surprising if we also consider that Morocco achieved a 20% real
depreciation during the eighties.
In this paper we try to uncover the reasons underlying the
performance oF the Moroccan economy. To anticipate our concluslons,
we argue that wage moderation and judiclous monetary pollcies were
instrumental in restraining inflation. Wlth one brief exceptlon ln 1983,
monetary authorities remained firmly committed to eschew any



2
inflationary financing of the budget deficit. This strategy could only
succeed however because of the wide ranging system of credit and
monetary regulations which worked to channel domestic funds toward the
Treasury at relatively lov/ coEt.-. The prospects for the continuation of
such strategy are not favou-able however. As far as the growth
performance is concerned, it Eppears that it can be attributed to an
outstanding export response to the new trade regime on the one hand and
a set of favourable supply, s.ucks, including  a string  of record
agricultural harvests and the collapse of real oil prices, on the other.
The organization of the paper is as follows. In the next section, we
study the evolution of the budget and of its different components. We also
examine how the deficit was financed. In section 3, we argue that the
reluctance by Morocco's policy makers to monetize existing budget
deficits is well explained by the sharply unfavourable trade-offs between
higher monetization and inflation that exist in Morocco. Such
conclusions are based on the estimation of a system of assets demand.
In the following section, we analyse the implications that continuing
budgetary disequilibria entail on investment and saving decisions. We
find that such implications may be substantial, even though they may not
work their way exclusively through traditional interest rates channels.
Frinally in section 5, we assemble the various pieces of econometric
evidence collected in the paper to study the impact of fiscal policy in
the context of a macroeconometric mGdel. The last section offers some
conclusions.



3
2. THE BUDGET DEFICIT: EVOLUTION AND FINANCING
a) the overaUl trend
The budget situation in Morocco registers a continuous deterioration
during the second half of the seventies. Sorme timid attempts to
macroeconomic stabilization provide only temporary restraint to the
burgeoning financial needs of the Treasury. However at the beginning of
the eighties the fiscal repercussions of the fali in Morocco's terms of
trade, the increasing burden of foreign debt attendart on the steep rise
oF International interest rates and the growing weariness of foreign
commercial banks to provide continuing financing' to the Treasury bring
to the forefront the issue of fiscal responsabilit y. As a matter of fact
the budget deficit, measured on a cash-basis, shows a slow but steady
decline starting in 1981. After increasing, as a percent of GDP, from
an average value of 3.2%  in 1971-1974 to 153% in 1976-1981, it
declines to 8% in 1984-5 and appears to stabi]ize around 5% in the
more recent years (Table 1).
Cash-based measures of the budget deficit can be however quite
misleading.  The Treasury may indeed resort to the use of financial
arrears to cope with mounting financial difficulties. This Is indeed what
happened in Morocco until 1985. Because cash-based deficits do .ot
allow for the accumulation of financial arrears, they will fail to refle
the full pressure that fiscal policy exerts on available resources.
Conmersely, cash-based deficits will overestimate the size of the fiscal
problem whenever financial arrears are being decumulated. Notlce in
this respect that, after 1985, adjustment programs in Morocco have
involved a sizeable reduction In the stock of a-rears. Overall, It Is more
appropriate to treat the accumulation of arrears as a source of
(involuntary) finance and measure the deficit on a payment order basis.
If this is done for Morocco, the deficit plcture changes substantlally.
Both the initial increase and the subsequent decline in the budget deflcit
are more substantial when measured on a payment-order basis than on a
cash basis (Table 2). This is simply due to the fact that arrears had
been accumulated until 1985 and decumulated afterward.
A further correction relates to interest payments. As column 3 of
Table 2 shows, the latter have steadily increased as a share of GDP,
compounding therefore the budget difficulties. Yet allowance should be



4
made for the fact that, under infltionary conditlons, a pote, allly
conspicuous share of Interest payments represents early amortiz, on of
outstandlng debt and, as such, should be considered as a (riegative)
financing component. For Morocco, historlcally low Inflation rates mean
that thls correction is not going to produce dramatic effects. Its impact
Is nonetheless substantial, albeit declining. From column 4 of Table 2,
we can see that capital gaIns on domestic debt due to Inflation were
equal to 3.1% of GDP in 1984. Due to the drop In Inflatlon, they then
declined to 1.3% In 1988. As a Xesult the operatlonal budget, whlch
only includes real interest payments, reglstered a relatively modest
deficit In 198x and a surplus in 1986. We have also corrected for
valuation effects In the stock of foreign debts, by multiplying the
outstanding stock at the beginning of the perlod by the excess of
domestic Inflation (measured by the change In the GDP deflator) over
the rate of devaluatlon (taken with respect to the nomlnal effectlve
exchange rate). We feel much less confident about this correctlon
insofar it overlooks valuatlon changes due for Instance to cross currency
fluctuations which appear to have played a substantlal role in
determlning the evolution of foreign debt indlcators for Morocco. We
nonetheless report the results of such correction (column 5). Its Impact
Is baslcally Inslgnlficant in 1984, but grows over tlme as domestlc
Inflatlc l is no longer matched by a corresponding devaluation.
The last measure we conslder is the primary deficit, which excludes
all Interest payments. By excluding a component whlch Is to a large
extent beyond the control of flscal authorlties, It provldes s more
accurate Indicator of the effort to redress exlstlng fiscal Imbalances.
The evolutlon of the prlmary deficit hlghlights the adjustment effort on
the flscal front. The primary deficit was equal to 7.2% of GDP in 1983
a.nd has steadlly declined since then to reach a surplus of 1.9% of GOP
In 1988. This amounts to a turnaround equal to nine percentage polnts
of GDP. The Improvement In the overall and In the operational deficlts
are less signlflcant due to the Increased burden of respectlvely nomlnal
and real Interest payments.
Summing up, it Is undeniable that Morocco has taken a declslve step
on the road to redress its fiscal Imbalances. The substantlal
achievements In the overall budget reflect an even more slgnlflcant
Improvement In the primary deficit whlch more than compensated the
rlslng burden of nomlnal Interest payments. The favourable evolutlon of



5
the budget was also facilitated by low and even negatlve real Interest
rates on domestlc debt. Thls situatlon Is however comlng to an end and
the Treasury Is now Forced to offer positive and hlgh real rates oF
return to convince domestlc Investors to finance Its deflclt.  The
average cost of servlilng domestic debt Is likely to rlse qulte rapldly In
the near Future, underscorlng the need for contlnulng flscal restralnt.
b) the primary deficit.
In this section we shall examine how the slzeable reductlon In the
primary deficit was achleved. The relevant data are contained In Table
3. It is apparent how the burden of adjustment fell mostly on publlc
Investment which registered a major fall fron  11.54% of GOP at the
onset of the adjustment program In 1982 to only 3.61% In 1986 and
recovered then, but only marginally, to around 6i In 1987 and 1988.
The drop In public investment was therefore the major contrlbutlng
factor In the process of fiscal retrenchment.  Thls Is an unfortunate
feature of the adjustment proc'ess that Morocco shares wlth many other
developing countries. We shall find in a later sectlon how the decline In
publlc Investment may have negatively affected the prlvate sector
propensity to Invest. Table 3 shows how other Items also provlded a
sizeable contribution to flscal restraint.  Expenditure on goods and
services fell somewhat, reflecting mostly the drop in public employees
real wages and the more sober trend in public employment. Simple
econometric analysis (not reported here) shows that expendlture on
goods and services is not related to It  ':on, suggestlng therefore that
the latter was not paramount in reduL- ig public sector real wages.
Finally we find that also the share of subsldies In GDP fell
substantially mainly because the decline in Imported food prlces
reduced the need for government Interventlo. . This Is Incldentally an
interesting example of the direct impact of terms of trade fluctuatlons
on the budget.
A comparison between the evolution of the primary deficit, as
reported in Table 2, and the behaviour of public expenditure in Table 3
already suggests that taxation did not provide a noticeable cuntribution
to the Improvement In the budget. As a matter of fact taxation as a
percentage of GDP fluctuates around 21% (Table 4). Had It not been for
the windfall revenue attendant on the petroleum levy introduced In 1986,
fiscal revenues would have actually registered a major decline over the



6
perioA 1983-'88. There are several causes behind thls unsatisfactory
evolutio-i. On the one hand this was a time where, with t'ie support oF
the World Bank and the IMF, the Moroccan government Introduced some
far-reaching reforms in the system of both direct and indirect taxatlon.
For instance the introduction of VAT in 1986 Is expected to represent
in the medium-run a more efficient and more reliable source of revenue.
Due to implementation problems it was however accompanied upon its
Introduction by a revenue shortfall. Similarly the overhaul in the system
of direct taxation with the phasing out of a set of Income taxes
differentiated according to income sources and thelr consolidatlon Into -
unique tax Is not likely to contribute to the budget In the very short-run.
Other factors may also have helped to determine the evolution of tax
revenues. For Instance it is often argued that high inflation will reduce
the real value of tax revenues because of delays in revenues collectlon.
This effect is unlikely to be significant in Morocco because of relatlvelv
low inFlation levels and may also have been offset by the bracket creeo
mechanism where taxpayers in an unlndexed system are taxed aL
increasing rates because of the Impact of inflation on nomlnal inrnomes.
Agaln econometric evidence does not suggest a slgnificant relatlon
between the share of tax revenues in GDP and the level of Inflation.
A further noticeable factor which affected the evolution of
government revenues was the process of trade reform. In 1984, with
the support of the World Bank, the Morocco government, in an attempt
to rationalise the trade regime and reduce the protection to import-
substituting productions, chose to substantially reduce Import duties and
gradually phase out the so-called special import tax, an across the
board tariff.  The revenue shortfall attendant upon this measure was
estlmated  to reach  4%  of GDP. Continuing budgetary difficulties
prompted a reassessment of the sitjation. In 1987 the special import
tariff had its rate inlcreased to 12.5%, from 5%, and its name changed
(to fiscal import duty). This policy shift is likely to lead to a growing
contribution of trade taxes to government revenues.
The last column of Table 4 highlights the impact of terms of trade
fluctuations on fiscal revenues. In Morocco production and trade of
phosphate products (the main foreign exchange earners for the country)
are controlled by the Office Cherifien des Phosphates (OCP), a public
eneterprise. The latter contributes to the Treasury budget through tax
and dividend payments. As shown in the Table, these contributions



-6a-
TA3LE 1
BUDGET DEFIrCT
(cash-basis, before debt  relief, ir, perzent of GDF)
Years                      Years
19''         2.9;          198:         13.64
1972         4.01          1982          9.22
1973         2.C:          1983         11.51
19I4         3.9           1984          8.07
1975         9.5'          1985          8.57
1976        12.C'          1986          5.69
197,        15.84          1987          6.11
1978        11.28          1988          5.45
1979        15.1-
1S8^         9.0'
Scurce: World Ba-.K data base.
TABLE 2
BUDGET DEFICITS: VAR-CUS DEFINITIONS
(as a share of GDP)
cash basis    payment order interest capital gains
basis       pavments  cn dom. debt
1983               0.115          0.12I          0.049           na
1984               0.081          C.112          0.061          0.031
1985               0.086          0.096          0.062          0.030
1986               0.057          0.053          0.059          0.035
1987               0.061          0.057          0.059          0.022
1988               0.055          0.041          0.060          0.013
capital gains       operational    primary
on foreig.n debt      deficit        deficit
1983                 na             na           0.072
1984               0.005          0.081          0.051
198'               0.019          0.066          0.034
1986               0.017          0.018        -0.006
1987               0.034          0.035         -0.00l2
1988               0.045          0.027         -0.0'.)
Source: 'orld Bar. data base
TABLE 3
CENTRAL GOVERNMENT EXPENDITURE
(before debt relief, in percent of GDP)
Years    Current  Capital  G & S    Interests  Subsidies
1971      14.12      3.72     12.51      1.01        0.60
1974      19.25      5.77     12.20      0.85        6.21
1977      18.58     18.93    14.44       1.50        2.63
1979      19.46     12.89     15.05      2.19        2.21
1981      24.60     12.33     15.42      3.88        5.30
1983      23.72      9.64    15.05       4.86        3.81
1984      24.00      8.09     13.63      6.08        4.29
1985      23.09      7.18     12.63      6.23        4.23
1986      20.62      3.61     12.20    5.92          2.50
1987      20.36      6.03     12.64    5.91          1.81
1983      2^.3:      6.34    12.17    6.01           2.13
Sc_rce: Worit Ear k da:a tese.



7
r-earhed a p-ak in 197/1 Vi, -,-n-onoitantly v. ith the phcsphate. pr.re
bo,^om. ThP, hav,e been declining since (except for a small rebound in
19B 1), adding therefore substantilc,v to budgetary diffirulties.
e) flwru:n 21 the budget dlefiit
The impact of budgst deficits on macroeconorm,ie conditions is to a
s.gniFicant extent a function of the mode of financing of the deficit Itself.
It is therefore essential to take a closer look at the way budget
imbalances have : een financed in " e past. Prior to 1983, foreign
borrowing flnancea nearly 60% of the Treasury deficit. The avallability
of foreign Finance came to Fin abruot halt in 1983, precipitating a major
payment crisis. The government was Forced to rely to an unprecedented
extent on Central Bank borrowing during 1983 and, to a lesser extent,
1984. This was also the period when, as it was mentioned earlier, the
government increasingly resorted to the accumulation of arrears as a
source of involuntary finance from the private sector. Yet the increased
monetization of the deficit would have unavoidably provided fuel for
inflation. It was the firm  commitment by monetary authorities to
eschew inflationary financing which prompted a major revision in the
financing strategy. Starting in 1984 the government increasingly began
tapping non-inflationary domestic sources of financing. The Treasury
reliance on voluntary domestic lending was enhanced by the need to
reduce the sizeable stock of financial arrears. This strategy was helped
by .he partial liberalization of domestic interest rates which was
associated with a major shift in the private sector portfolio
composition from  currency to time deposits. At the same time the
ceiling on credit to the economy forced commercial banks to channel a
substantial part of these financial resources to the Treasury, creating
thereby a steady source of finance for government deficits. Another
important component of the financing strategy was the direct sales of
Treasury bills and bonds to the non-bank sector.
The counterpart of this strategy was its increased cost to the
Treasury. To make time deposits more palatable to commercial banks,
they were excluded from the base on whict obligatory investment in
Treasury bonds are calculated. Together with the reduction of arrears
and the need to offer att  tive (post-tax) returns on direct issues of
government bonds, this measure induced a significant increase in the
cost of domestic financing for the Treasury. This is -Iearly reflected



in the rapid increase in the average real Interest rate on government
domest ic  debt,  as  noticed  in  the  previous  paragraph.  Further
iristitutional developments arz likely to reinforce this trend, iF existing
pr-oposals to reduce liquidity and portfolio requirements for tne banking
sector are implemented. In the last section of this paper, we shall
assess the macroeconomic impact of furthering the process of financial
liberalization.
We can try to sum up the main conclusions of this section and draw
some implications for the future course of fiscal policy in Morocco.
Overall Morocco has been quite successful in coping with VormIdable
financial difficulties. The foreclosing of foreign borrowing In 1983
represented a major shock to the economy and to the Treasury, more
significant perhaps than the rise in international Interest rates.  The
two-prong response of M4croccan policy-makers relied on a sharp
reduction in the Treasury i,nancing needs and a shift in the composition
of finance with a view of eschewing inflationary pressures. A key to the
success of this strategy was the availability of relatively cheap sourcs,
of domestic finance, mostly determined by a complex system ,t
Financial regulations. However the difficulties of relying at the margin
on financial repression as a cheap source of funds, together with the
relaxing of fiscal discip,ine in the most recent years, have already
prompted a reassessment of ex'sting strategies and, perhaps merea
crucially, highlighted the shiarp trade-offs that continuing budget
disequilibria entail. Today the major question facing Morocco's policy-
makers is 'shether the fiscal policy stance is consistent with the
mainterance of low inflation, the resumption of investment and growth
and the external payments constraint.
Consider the not too unlikely case where foreign finance will not
increase substantially in the medium-run. Under these circumstances,
the failure to oersevere on the road to fiscal discipline may entail
severe macroeconomic consequences. Even the perpetuation of present
financing strategies of a relatively unchanged budget deficit would soon
run into severe problems. First of all the bucget is still evtremely
vulnerable to international interest rates and terms of trade shocks.
For instance the revenue from the petroleum levy which in 1988 still
represented 3.4% oF GDP is likely to vanish under increasing oil prices.
Second the cost of domestic finance is likely to rise steeply in the



9
future. Again the Treasury budget would be extremely vulnerable to such
evolution. Would the cost oF servicing domestic debt Increase to
competitive market rates, this would add to the budget an extra burden
equal to around 2% of CDP.
Could Morocco resort at the margin on larger monetlzation of the
deficit ?  By internatiou,al standards Morocco s inflation Is typlcally
low. An increase in the Inflatlon tax may then represent a palatable
alternative. However, besides the danger of tampering with monetary
policy and damaging the credibility of the Central Bank commitment
against Inflation, thero. Is ac-tually little room  for a substantial
contribution to budget Financlng from Increased monetizatlon of the
deficit. We will show in the next sectlon that, for Morocco, estimated
elasticities From a system of assets demand indicate an extremely
unfavourable trade-off between inflation and monetary financing.
Continuing reliance on non-inflationary sources of domestlc finance is
thereFore essential to keep Inflation under check. However this strategy
could result in large increases In domestic real rates of interest.
Besides putting into jeopardy budget equilibria, this would most likely
crowd out investment. Moreover even if the increase In interest rates
did not materialize and the perpotuation of a system of financlal
regulations guaranteed a source of inexpensive finance for the Treasury,
the impact of unabated budget deficits on investment and on growth
would still be signi ficant because of their effects on credit markets.
We shall examine this issue in section 4.



-9a-
TABLE 4
TAX REVEENJES (as a share of GDP)
Total         Income   Taxes on goods Taxes on
revenue         taxes       and serv.  int. trade
1971               0.1'849        0.C282?        0.07604       0.02582
1972               0.14401        0.02814        0.07402       0.02407
1973               0.16157        0.03206        0.07454       0.02879
1974               0.21112        0.03294       0.06797        0.03625
1975               0.23329        0.07023       0.07320        0.04331
1976               0.20292        0.04011       0.07688        0.04184
1977               0.21672        0.05145        0.08326       0.05187
1978               0.2120i        0.05343        0.08012       0.04993
1979               0.22246        0.05364        0.07990       0.05477
1980               0.20506        0.04507       0.07836        0.05562
1981               0.22570        0.04765        0.07642       0.06190
1982               0.22046        0.04125        0.08410       0.06235
1983               0.21272        0.04380        0.08768       0.05307
1984               0.2v890        0.04438        0.08510       0.05009
1985               0.20653        0.04337        0.08376       0.04456
1986               0.18891        0.04040        0.07089       0.03629
1987               0.20679        0.0.475        0.07656       0.03561
1988               0.22582        0.34717       0.07793        0.03954
petroleum         other           OCP
levy          taxes        contributions
1971               0.00000        0.01836       0.00182
1972               0.00000        0.01778       0.00176
1973               0.00003        0.02617       0.00609
1974               0.00000        0.07395       0.06306
1975               0.00000        0.04655       0.06776
1976               0.00000        0.04408       0.02316
1977               0.00000        0.03014       0.01843
1978               0.00000        0.02852       0.01450
1979               0.00000        0.03415       0.01128
1980               0.00000        0.02601       0.01080
1981               0.00000        0.03973       0.01641
1982               0.00000        0.03276       0.00581
1983               0.00000        0.02817       0.00756
1984               0.00000        0.02933       0.00979
1985               0.00000        0.03484       0.01342
1986               0.02444        0.01689       0.00000
1987               0.02782        0.02204       0.00290
1988               0.03375        0.02743       0.00401
Source: World Bank data base
TABLE 5
IMPACT ON INFLATION CF A 1% INCREASE IN THE GDP SHARE
OF MONETARY FINANCING
Initial inflation      low elasticities               High elasticities
4%                    7.5%                           7.6%
6%                    9.8%                          11.1%
8%                   12.9%                          16.9%
10%                   16.7%                          26.4%
Source: own calculaticns



10
3. ASSETS DEMAND, SEI-;INORAGE AND THE INFLATION TAX.
Received wisdom argues that, following the drying up of commercial
lending after 1982, developing countries had to make a twin transfer of
resources: one abroad to foreign creditors insofar as new lending had
fallen much below the required service on outstanding external debt; the
other internal, toward the governnent, to the extent that in several
developing countries the State had taken onto itself the burden of
servicing foreign debt. Together with the increase in interest rates and
the fall in terms of trade, this has often meant the disruption of
budgetary equilibria, already jeopardised in many cases by previous
unsustainable fiscal policies. The inflexibility of the tax system, the
downward rigiditv of fiscal expenditure (with the noticeable exception of
public investment) and the thinnes of domestic financial markets left
then little choice to local policy-makers but to monetize the fiscal
deficits with sometlmes calamitous consequences on the inflation rate.
Under this interpretation, the exceptional inflation record in Morocco
is undoubtedly puzzling. In the previous secticn it was shown how
financial repression, by channeling low cost funds to the Treasury, was
instrumental in a strategy of relying on non-inflationary sources oF
budegt financing. Yet the question remains of why Morocco's pollcy-
makers remain staunchly opposed to even a limited monetlzatlon of the
deficit and a greater reliance on the inflation tax. As a matter of fact
inflation in Morocco is extremely low, 2.3% in 1988, even by the most
stringent international standards. A recent report (United Nations-
World Bank, 1990), wh le not explictly advocating an increase In
Inflation, argues nonetheless that there is some mileage for boosting
revenues from the inflation tax. The issue however cannot be solved on
purely theoretical grounds as the inflationary implications of deficlt
monetization depend on the response of money base demand to changes
in inflation and interest rates. In what follows we shall argue that the
outlook for greater monetization of the budget deficlt In Morocco Is
altogether unfavourable. If we also consider the destabilising effects on
Morocco's social fabrics and the loss of Central Bank credlbility that a
higher inflation would Imply, we can perhaps understand the firm
commitment of the country's policy-makers in favour of prlce stability.
The scope for increased government revenue from seignorage and the



11
inFlation tat< is determined by the private sector's choice of assets. In
this paragraph we rely on a standard Tobin's portfolio approach to
analyse the demand for currency, demand and time deposits. Other
assets are not included in our analysis on the ground that, with the
exception of real assets, they play a minor role in the private sector's
choices. In what follows we assume that the demand for each asset Is
related to its own return, the other assets' returns and the level of
income. As a proxy for the return on real assets we use the (expected)
inflation rate, which however is not observable. We assume that
expectations are formed rationally and depend on the set of inFormatlon
available at time t-i, which is defined to include the lagged level of
prices, the money supply, the wage rate and the exchange rate.
According to our estimates, the last variable does not contribute
significantly to the prediction of future prices. This prellminary
evidence suggests that, for Morocco, the pass-through from the nominal
exchange rate to domestic prices is Fairly weak accounting therefore for
the fact that a sizeable real depreciation did not translate into higher
inflation. Similarly the significant impact of wages on prices supports
the claim that wage moderation (the real minimum wage declined by
7.3% from 1983 to 1988, after increasing by 17% in the three prevlous
years) was instrumental in containing inflation.
The fitted value from the price equation is used as an estimate for
expected inflation. In the process of estimation, we rely on Pagan's
(1984) procedure and use our proxy for expected inflation as an
instrument for actual inflation. This procedure is designed to yield
consistent standard errors for the coefficients. We still have to
determine whether the system of assets demand should be expressed In
nominal terms (with prices included among the explanatory variables
and the relevant homogeneity assumptions tested) or in real terms. We
take the first course for the sake of generality and assume thereby that
assets demand do not fully respond instantaneously to changes in prices.
We also allow for the possibility of lagged adjustr1ient to income and
interest rates variations. Therefore the estimating equation for a
generic asset MW reads as:
In MJt  ao + a, In Pt + a2 In Pt_l + a3 In Yt + a4 In Yt  + as it
+ a6 it1 + a7 Tre + ae In M(I)
t-i  t ~ Mt-i



12
where p and Y denote the prlce and Income level respectively, "I"
represents the vector of assets' returns and Tre Is expected Inflatlon. We
would expect that in the long-run the prlce elastlclty of assets demand is
equal to one, I.e. that assets demands are homogeneous of degree one in
prices.  Formally this implies t,Lat a1+a2=i-a8. We can reparametrlze
the prevlous equation to test this restrict3on:
In (Mj/p)t = ao + (a,-1) In Pt + a3 ln Yt + 84 ln Yt  + as It
+ a6 i   + a7 wet + (a8+a2) In pt_, + a8 ln (MJ/p)t_1  (2)
If long-run price homogeneity holds, we have
ln (Mj/p)t = ao - (a2+a8) A In Pt + a3 In Yt + 84 In Yt  + as it
+ a(, 1 + a7 wre + a8 ln (M /P)t    (3)
Eq. (3) shows that Imposing long-run prlce homogenelty In a context
where assets are expressed in real terms adds a backward-looklng
Inflation term, A In p, to the equation itself. This term however does
not reflect any substitution effects between real and financlal assets, but
only lagged adjustment to price changes. Traditional money demand
equations where Inflationary expectations are simply modelled in a
backward-looking  fashion  may  therefore  mistakenly  Interpret the
significance of A In p as evidence of substitution toward real assets,
whereas only a process of dynamic adjustment Is Involved. Our
approach allows to distinguish between these two effects by separately
evaluating the statistical signlficance of A In p and Tr . Admittedly
though, the coexistence of these two effects may be difficult to justlfy on
purely theoretical grounds.
In the estimatlon process we include in the tector of assets' returns
("i") only the Interest rate on time depo_its, I . Some demand deposits
are also remunerated, but their return moves in a almost perfectly
collinear way with the one on time deposits. We have therefore Included
the remunerated component of demand deposlts in the aggregate of time
deposits. We can now describe the results of the estimatlon of the
system of assets demand. Our estimation strategy Is as follows. We
start from the more general dynamic specification In eq. 3 and restrlct
the model to obtain a parsimonious representation of the data generatlng



1 3
process. In what foll-ws ,e only report the F.nal e,quation.
All equations are estimated over the 1974-1988 period.  The last
year however is saved to test For out-of-sample stability. We begin by
the demand for currency.
A In (CUR /pt)  -. -1.10 itd + .30 In Y - .60 In (Pt/Pt_1)
t t  ( 17i (. 17t       (.07)      (.14)
- .70 In (CURt iPt- l)
DW = 2.01  SER   .013  LM - .25  Hendry = 1.47
The demand for currency  (CC9E) is significantly related to the
interest rate on time deposits (i ) and to income. Actual inflation
appears in the equation with a short-run elasticity of .60. This is not
necessarily,  as  mentioned  earlier,  an  indication  of significant
substitution possibilities between currency and real assets, but may
simply reflect the fact that nominal demand for currency fully adjusts
to changes in the price level only in the long-run. The coefficient on
expected inflation instead is not signiFicantly difFerent from zero and
bears the wrong (positive) sign.  Jointly estimating the price and the
currency equations (and testing for the expectational restrictions) did
not improve the results. The variable 7re therefore has been eliminated
from the final equation. Finally the hypotheses that nominal demand for
currency be unit-elastic in the long run with respect to the price level
and to income are not rejected by the data (tio = .94). As diagnostic
tools we rely on the Hendry test for out-oF-sample stability and the
Lagrange multiplier test for serial correlation. Both are distributed as
X (1). They do not provide any indications of misspecification.
We now turn to the equation for demand deposits (DD).
In (DD /Pt) = -11.65 - 8.46 i td  1.03 ln (pt/pt1) + 1.20 ln Yt +
t  t   (4.26) (2.96) t    41)                (.37)
.76 ln Y
(.26)
DW = 1.43  SER. = .033  LM = 1.56  Hendry = .81
Demand deposits have been defined to exclude saving deposits.



14
Following this modification, we find that both inflation and the return
on time deposits pla>E a significant role in affecting demand deposits.
The long-run income elasticity is equal to 1.96 and significantly
different from one. Once again expected inflation does not contr'bute
in a statistically significant way to the equation. There is no sign of
misspecification as indicated by the usual battery of tests.
The last component of our menu oF assets are time deposits (TD):
A In (TD /Pt) = -1.54 + 6.94 I td + .96 [(In Y - In (TD[  /Pt 1)]
t t    (.8D) (4.22)t    (.18)    t          t- t
DW = 1.97   SER = .09  LM = .02  Hendry = .18
As expected, time deposits respond positively, albeit not very
significantly, to an increase in their own rate oF return. The dynamic
specification of the equation is very simple. There is no evidence of
lagged adjustment to prices as indicated by the insignificant coefficient
on actual inflation. On the same ground, also expected inflation has been
excluded from the final equation. It was not possible to reject the
hypothesis that the demand for time deposits Is unit-elastic with
respect to income (F1,9 = .54). The LM and Hendry tests Fail to point
to some misspecification problems.
Overall the previous results indicate that the demand for monetary
assets In Morocco is strongly influenced by the patterns oF returns. The
estimated semi-elastlcltles on the interest rate on time-deposits suggest
potentially conspicuous shifts in portfolio composition In response to
variation of the structure of interest rates. There is however less
Indication of strong substitution posslbilltles with respect to real
assets. Our proxy for expected inflation never proved to be significant In
any of the three equations. It is not apparently a problem of
multicollinearity insofar as the coefficient on actual inflation was quite
well determined even in the more general specifications. It is not either
a problem of statistical methodology. Our approach should provide at
least consistent estimates oF the coefficients and of their standard
errors. As mentioned earlier, more eFficient simultaneous estimation
methods which also allow for the expectational restrictions fall to
change the basic findings. Perhaps our proxy for expected lnflatlon Is
not well specified. Alternatively, expectations may to a large extent



15
have an adaptive form. We leave the matter unsettled.
To c,-mpute the amount oF monetary financing the government can
count upon, we must allow for the fact that in Morocco a distlnction
must be made between reserve requirements, which apply only to
demand deposits, and liquidity requirements, which force commercial
banks t,- invest a fixed share oF their deposits in low yield treasury
bills. Recent financial sector reForms have however substantlally
inc reased the interest rate paid on liquidity requirements which now
approaches market rates levels. We include therefore liquldity
requirements in the domestic public debt. The narrowly defined
monetary base (MB) is thereFore equal to:
MBS, = c t + rrdd DDt                                      (4)
where rrdd is the reserve requirement coeFficient for demand deposits.
At any point of time the amount oF monetary financing is equal to the
change in the monetary base. As a share of nominal GDP, the amount of
monetary Financing, A MB t/(pt Yt) is equal to the rate of change of
monetary base, (A MBt/MBt), times the GDP share of MB, (MBt/ptYt).
We can now use eq. 4, together with our estimates of the demand of
currency and demand deposits, to evaluate the relationship between
inflation and the amount oF monetary financing. We focus on a steady
state situation and arbitrarily impose that demand deposits be unit-
elastic with respect to income. As a result, the growth rate of MB will
be equal to the sum of output growth (n) and inflation (Tr). Similarly
the GDP share of MB will depend on the level of nominal interest rates
and, possibly. of inflation:
CUR   rrdd D               r d
A MB/ (pY) =  +n) CUR + rrddDD  = (Tr+n) [ A eacr+ rrdd B e 9 ]
p Y                               (5)
where the coefficients A and a (B and 6) are computed from  our
estimates of the currency (demand deposits) equation. In assessing the
value oF a and 6 we face a basic ambiguity. Changes in the inflation rate
can have a direct impact on the demand for monetary base to the extent
that they lead to cor.-esponding variations in nominal interest rate. We
assume this to be fully the case. But changes in Tr have also a direct
effect on the demand for currency and deposits. However if we interpret



16
the signiFicant coefficient of actual inflation in our estimates as simply
evidence of lagged adjustment, then this effect should not reasonably
play a role in this steady state kitid of analysls. If alternatlvely we
believe that expectations are adaptive and our results reflect the
existence of significant substitution possibilities with real assets, then
inflation should have an indipendent effect. Notice that, In the latter
case, changes in inflation will exert a larger Impact on the demand for
monetary base. In what follows we allow for both posslbllltles.
We can use eq. 5 to compute the effect of a one percent change In the
inflation rate on the quantity of monetary financing as a share of GDP.
We take GDP growth to be equal to 4 percent. From eq. 5, it Is clear
that this derivative is a function of the Initial share In GDP of currency
and time deposits, which in turn depends on the Inflation rate Itself. We
can also compute the inverse derivative, which measures the Ir-rease
in the inflation rate attendant on a one percent Increase in the GDP
share of monetary financing (Table 5). Our estimates suggest that the
inflationary impact of higher monetization Increases very rapldly and
the trade-offs between Inflation and inflation tax worsens zubstantially
as inflation increases. This provides considerable support to the cholce
of Moroccan policy-makers not to rely on inflationary forms of deficit
finance.



17
lN. I\L'KTM-'N  AN' `AVINi A    'I LLIIIDN
AN h'sth rat-~ -I in.-~?tm;nr--r e-.ents a basic condit'ion fo-r ru-taine'
!n.x-agc~  in ec-nomn,- gr-oxth -,e  the long-term. Since 1992 ho\ever
the fGUrre-nt Fpicer  'p  chare of invostment in Moro-cco has been
stoa1i'. falling %,ith the o nl, e<-eption of 19_RS'. raising increasing
c2Dncer nr about th"- long-run perFs-pecives oF the economy. L,oo.king at tho
>:-nstant price ratio bet.v.Feen int-stment and iGP wouldi onlP  accentuate
tr.- Fall' in in .est`ien, beause  .' the irncrease in the real price of
in. estmentF goods .attesndant on the real deprecdation. More wsiorry,ingly
porhaps, the drop ir .acital a-cuTnlation is a generalised cne, in.dylving
bot,h pub:i> andr pr-i'ate in.estment. Contra-v to initial e<pectati':ns, the
fi-a1' rtrenchment v<.h;,ch to-k a toll or public investment was not
- rnsa -nEtE,d  , a matc-hing irr.zrac-- in private investment. Is the Fall in
:n'. eatmont ailEnlficart cause F,or -c,ncern 7  Surely enough, it co-uld be
a-rgue-d that m '-t :)F the Fa'l in .rEvestment can be predicated, on the
high^er c,-st    epr cap tal.      enco uaging  less capitalintensiv,e projects
(ardi remo>Xving the pre.icus Fbias a&falnst labourintensive productions:', an
i-crease in the      of c-apital wou  allow the same level of grow>th to
be  ae;ed    ith  a  l. er  volume  oF  investment.   LUnder  this
nt-rpretIatln thD  ropi in inestunt would on.;, represent the out-ome
F atcusting to a no.. constellatioDn of factor prices. Empirical evidence
-r a roJatiel>  large samFle co  developing countries (Faini and de
!-:-. 199s  rufLggesto  that CcsDt ,-F capital considerations can account
on'. f-r a small fraction of the fall in investment and other factors
t!he-ref-re must be- at .%vork. In  ,,,hat follows:, We shall assess the
r>-.-s,>- F9--f thlis aFproach for the --ase of Morocco.
We- ro   -r.  a smzer   m.de' rf a Firm. The fir,m  is assumed to
a  m:<1-!7e itS rne-t v-. -rtk. sub c--  1- a standard nsoolasK' -ai p'-oluctio-n
r--t.- r-cr.t.-a-t  . tl- the  itionalt s-t-up hov-.e,vr. financ,ial
a--sl   '   h    a shFi -'iant boearing on real d       ' E
o mr-'.-<' th>e 9.mQ>a-t I f 9nancia, variables by assuming that e<ternal
-suit`. fiErancing is una,.ilaHle and the firm  must rely on two alternative
c-.urmzcs oF in.-stmo-nt F1narde, i.e. retained earnings and bank's debt.
von that the entreproeu rs' discount rate is assumed toD) be larger than
the  rik-Fr-ee  int-reFt rate  (otherwise  the  Firm  would  accumulate



18
fr.naa % acs.tt . doLt; is', theF favoured ErUri oef in\eswtmcnt f:nan,re.
prl;>.- Ferhap  tF au-- .f ta   :nsiderat ios).  Ar internal sciin fc:r
the optimal dett choice of the firm  still! e<1-ts. though. prnxbided that
v.e assume that h!gher debt, relativel',y tD the firm's capita! stor k, is
as,lt  '.52th increasing agenc, costs. Emal'l, because of -oQnstraints
in the  credit market, at each point of time the adebt to capital r.ticfk
rati,- of the firm is b.ound by an upper ronstraint. Formally,:
maK (           F. Y       Nt - qt It + Et   1itr)-1
1~~~~~~~~~~~ t-            C(t 'It Kt!]    (.
s.t.             K �  f*  qt Kt                                (7;
t   FKt.N                                  (8'
KCt  fI:iKs_, + It1c
'x.here. in standard notations, Y, N, K and I represent the level of
pr oduction, employment, capital stock and  investment respectively,
whereas B denotes the outstanding stock of debt. The function et ! with
c >0. cjj')  and c2<0, is the agency cost function, which is. assumed to
be quadratic.  The output price is indicated as p, the wage as w, the
price oF investment goods as q, while r and i denote the interest rate
and the discount rate respectively. We assume that I+i= (l+r) (I+k)
N.here k is a multiplicative risk premium. Finally d is the depreciation
rate.  Eq. 6 is the leverage constraint whi-ch defines the madmum
amount of debt as a time varying proportion (E*) cf the capital stock,
while eq. 7 and 8 describe the production relationships and the capital
accumulation  identity  respectively.  Suppose  first  that  the  debt
constra;nt is not binding. It can be easily shown that, at an optimum,
debt Vill be a fixed proportion , say y, of nominal capital stock.  In
turn  v is a function of the risk premium  k and oF the parameters
aFfecting the position of the agency cost function 2 Notice that y does
not depend on the interest rate. As matter of fact, variations in r (or in
the discount rate i) affect the level of investment, not the con.position of
its financing. The choice oF y in turn afFects the demand for capital
which is otherw.ise determined in a standard way. IF we consider the
case  where  the  debt constraint is binding  (i.e. B*<y), financial
conditions have a direct impact on the demand for capital. Under linear



1 9
'''-'r-lit.  ':n9>i:n''r  inr F:r--:i,ti.:n. it c-arn t: >, shown that the optimll
-> ta*. ¶t>'K  ..:'. i:.   n a furw't .-r of zutrut.-F  the r,aa  cost o-f capit-l rini
the Aa.la liti- --f d4t:t. FsrniaJiv:
KC   K  Y . -4
i (1                          p 4-  I  arameters r and p den.:te the
:--rprt>  ta  rate and the percentage reduction  in q indiuc-  t. the
s   -   ,r fiscal and  financial in-entives available t,  investors in
Mi.-r~-r-.  Fu'l details For the calculations of r arnd p are provilr-l in
W-Drlid dlk (1990?.
fJne -r'2,blem  .w;ith the Formulatio-n of eq. 10 is that it containr one
oh' i-uF']  Ssndigenous variable, the le,el of output Y . In what follows we
rel; lr th- Following procDedure. It is assumed that, because oF delivery
lags, flrms must determine their desired level of output, and therefore
their investment, one period in advance. They will need therefore to
predict, based on available inFormation, the optimal level of capaoity
outrut For the Following period. This will in turn determine their
demand fo r investmenrt gcods. We alsco assume that expectations about
thp determinants of the output decision can be simply modelled by a
Fir-st-o-rder autoregressive process. The expected, optimal level of cutput
(t- Iy t ' ,,iJll thereFoDre be equal to:
t-' IYt =Y (Ft- I , '>t- I,'t _t-I  PUbt_ tv-1'st-1!            II
w.here the inFormatilon set has been augmented1 to include both Ipub (the
level cF public investnment on the ground that this may affect the
pro-ducti,on relationship  > and MS (the stock of money, as a further
predictor of prices). In estimating eq. I 1 we must allow  For the fact
that t Yt is not observatle. Howeve-, under rational expectations, it
will YVfetr From  actual output Y only for a random  term  uncorrelated
w\.ith an'. available informaticn at t-I. We can therefore use the actual
product :-n level Y  Es the dependent variable and take the Fitted '.alue oF
eq. I I as the estimate of expected output.
2'. estimating eq. 11, it is now possible to take into account the
endogeneity oF Y in eq. 10. We follow again Pagan's (198t suggestion
and take the estimated value of Y (Y.F) in eq. 1I as an instrument For
the actual level oF output in eq. 10. We present First the estimates of



20
~~(ri    .u(1:.                             i           1 
+C.: (I,    . Ye    +    C. +9Mo -      l
!     F-u'  k.K+t tim  s.C  di:.9
N. 'XV)   LV\! =  AX-.  WK:  - .r5   LNAm =.5
w,here- dil  iF a  dummy  variab-e  which  takes  a  '.r!ue  of  I  in
coDrrespcndanm2e of the agricultural negative supply shocks in  1981,
19.33   and  1937. The wage  rate  dc1es rot pro.tide  a si-ni6ant
ccntributicn to the equation. This may be attributed to- measuremenrt
errors (\We use an indicator of the minimum  wage) or to the fal-t that
labcur does represent a signifIcant ccnstrdint. T he cocst of c acpital
instead appears to plaY a more significant rcle together with public
inv,estment and the money supply. As a diagnostic t-ol, we relv on the
Lagirange Multiplier test for serial correlati:n, which does not indicate
any significant problems. Mloving> nov.' to the investment equation, we
find that:
A In lt -3 .555+ 1.9 4 A InY  - .6  In (l'Y   - - .7  In (e/F:
(q1.82)  ( .83)   t              '1           (c/pt-    Bt
5-.67 (2 /Y)
2
R  -.50  DW= 1.i1   SER P .12
whe;re I represents private investment. E /Y, the ratic oF firms credit
to ';Ci7, is used as a pro<y of the stringerFcv at an aggregate lev,el of the
debt constrain:. The equation hIs b-een estimiatd by an inst-umental
.ar abl9e pFrcedure, with the fitted value frorr eq. 11 as an instrunmonn
f-r Y. The restriction that investment be unit-elastic w,ith rspect tu
_utFut has Leen tested (tI2z .5) and imposed or the data.
The estimation  of eql. 10 yields two interesting  results.  First
investment is signiFicantly affected by the real cost of capital. Fiscal
policy can therefore affect the investment decision through its impact



21
cn interest rates or, more directly, by changing the set of fiscal and
finanrial  incentives  available  to  investors. Second  the  level of
InN estment in the economy will also depend on the availability oF credit.
The joint presence of both the cost of capltal and oF credit avallability
mak appear redundant (or even contradictory), until we recall that
investment in this model is not fully determined by the stock of credit
that financial intermediaries are willing to extend to firms. Even if
Firms are credit constrained, a change In the interest rate r wlll still
affect investment through its impact on the discount rate i and thus on
the choice cf retained earnings. By varying their retention behavlour
firms can relax somewhat the credit constraint. Fiscal policy therefore
will aFfect investment by influencing either the Interest rate or credit
availability. This latter channel, as we shall see later in the paper, can
play a crucial role in determining the macroeconomic outcome oF
different fiscal Folicies. It is also plausible of course to interpret ths
joint significance of credit and Interest rates in eq. 10 as implying that
only a subset of firms is credit constrained.   Finally, by putting
together the estimation results for eqs. 10 and 11 we can argue that
public investment bears a complementarity relationship with private
investment. Indeed an increase in public investment will lead to higher
capacity output (eq. 11 and, through this channel, to higher private
investment. By severely cutting public investment, fiscal policy may
have contributed in the past to the stagnation of private investment.
b) the saving decision
A steady supply oF domestic saving will ensure that a sustained rate of
investment would not be incompatible with existing constraints on the
external payments front. It is essential thereFore to gather an adequate
understanding oF the determinants oF saving behaviour. Unfortunately the
measurement of saving is beset with difficulties, to the extent that
consumption is computed residually in Morocco's national accounts. We
can rely on two alternative measures of private seving, one derived as
the difference between private disposable income and private
consumption, the other based on the saving-investment identity for the
economy. Because of statistical inconsistencies, the two procedures do
not yield the same result. In what follows we use the first measure. We
find that, after peaking during the phosphate price boom and declining
aFterward, the average propensity to save has been steadily increasing
during the 1980's.



2) L
'~ lthe behavcur cf sa \ing, we first take a simple permanent
acrroach. Under well-known ccnditons the maximization oF
inter4emporal utilit'y by the representative consumer %ill Imply that
aSgr-Fate ccrnsumption (denoted as C) is simnpl equal to a proportion of
permanent in-ornme (YR), i.e.:
C - k YP                                                     (12)
Note that C should be defined to inloude only the consumption of non-
durable commodities. Lack of data preclude this important refinement.
For estimation purposes, at least two issues need to be addressed. First
we must specify an indicator of permanent income. In what follows, we
simply regress the actual value of real disposable income on a time
trend and take the fitted value from such equation as an estimate of YP.
Secondjlv we need to recognize that the parameter k will not in general
be Ficed, but will depend on the real interest rate and, possibly, the real
exchange rate. The impact of both variables on the propensity to consume
is however theoretically ambiguous. Our estimated equation reads as:
A ln Ct =13 + .48 (I'n Yt - In C: 1) ~ 17 r - .044 In X
t  . 12) (.08)                (.10)   (.018)
R2 _73    W =2.25  SERP   .015
where r and X denote the real interest rate and the real exchange rate
respectihely. The latter is defined so that an increase in X implies a a
real depreciation. In the estimation, the restriction that the long-run
eiasticitv of consumption with respect to permanent income be equal to
cne has been tested and imposed in the equation (t14 = .81). Our
results indicate that an increase in the real interest rate and a real
depreciation will both lead to a decline in the propensity to consume.
This approach highlights two channels through which fiscal policy
wE. affect private saving and consumption behaviour. First, fiscal policy
ma. influence, through taxes and transFers, the volume of disposable
income the conaumers can spend. In this framework a temporary tax
increase will have a more limited eFfect than a permanent increase.
Second. fiscal policy may aFFect private sawing bv inFluencing the level
cF irnterest rates. Our estimates suggest that an expansionary fiscal



2 3
poIi'-', '4,~i. cr-,'~. iq ot thr :ush the intZFA rte- cha!nnel .L-th rcn!.-unmption
and in.estmPnt d-mand. A third -hann,e (n.t allowei for in the previous
estimatesi may be howe er as v ,-r  if rati-na! Fri'.ate agents takie fully
into a,-count the future ta< liatijities ass--ciated with bornd-financed
defic~it.  The impli-cati-n Is tha' a shift from ta< to debt finance of a
given volume of public e<enditure should be neutral insoFar as it would
be mat-hed bY offs`,tting priaete agents behaviour. This is because
private agents would be perFectly aware of the future tax liabilities and
the consequent reduction in their permanent disposable income that the
increased deficit entails and would correspondingly reduce their
consumption. This is known as the Ricardian equivalence hypothesis.
There are several reasons whv this pr oposition may not hold, among
which it is worth mentioning the possibility that private agents may
discount the future at a different rate than the government and the
presence of capital market imperfections which hinder the intertemporal
smoothing of consumption by private agents. Evidence on these issues
for developing countries is limited (see however Rossi, 1988, Nam
1989, Haque and Montie!, 1989, Deaton 1990 and Haque, 1990). In
what follows we amend the previous specification to allow for the
possibility that, by reducing perceived disposable income, government
deficits may have a negative impact on consumption. We do not expect
this effect to be particularly strong in Morocco, because of the absence
of a consumer credit market and the likely pervasiveness of liquidity
constraints on households. To capture this effect we introduce liquid
assets (currency plus bank deposits) in our equation. We also allow for
inflationary effects on the ground that a high level of inflation will lead
to capital losses on liquid assets and impart a downward bias to the
traditional measure of disposable income. An increase in inflation
should also be associated with greater uncertainty and lead, through this
channel, to an increase in precautionary saving. If we estimate this
more complete model, we find that:
A In Ct = .46 (In Y  - In C,_ ) + .06 (n Mt - In Ct) -.09 In X
(.056)               (.02)                 (.02)
-.12 A In pt - .13 (deF/(CODP)
(,07)        (.05)
R2= 83  DW = 2.16  SER  .012



24
V.hCFC. in standard notation, F indicates the price Ie el (so that A In p
appFroimates the inflation rate-, M1 denctes the stock of liquid assets and
depf/GDP  indicates the ratio of the budget deFicit to GDP. In this
specific-ation, which closely follcsv Hendry and von Ungern Stenberg
(1981), consumers adjust theIr expenditures to ensure constant steady
state equilibrium ratios of consumptiun to disposable income and to
liquid assets. These long-run equilibrium ratios are affected by the level
of the real exchange rate, by  inflation and by budget deficits. The
significant impact of budget deficits on consumption in Morocco may
seem surprising, but was already found in work by Montiel and Haque
(1989) and by Schmidt-Hebbel and Muller (1990). We have tried to
assess whether this efFect could be attributed to the fact that government
expenditure on goods and services substitutes for private consumption,
but the high collinearity between the deficit variable and government
expenditure prevented such a test. Finally both higher inflation and a
depreciating real real exchange rate have a negative impact on
consumption.  The real interest rate instead is no longer statistically
significant and has been excluded from the equation. We can only
speculate why the real interest rate no longer seems to be a significant
determinant of consumption behaviour. One possible explanation could
be that in the previous specification the real interest rate variable was
actually picking up the efFect oF budget deficits. Indeed, as our last
estimates suggest, a larger budget deficits lowers consumption and Is
also likely to be associated with rising interest rates. Overall this new
specification appears to indicate an even stronger role for fiscal policy
in influencing consumption behaviour.



25
5. MODELLINGi THE IMPACT OF FISCAL POLICY
A comprehensive analysis of fiscal policy requires an economy-wide
model. In this section, building on the estimates already presented and
on previous work by Faini, Porter and van Wijnber gen (1989), we
present a simple macroeconometrlc model which will then provide the
basis for an evaluation of fiscal policy In Morocco.
A full presentation oF the model goes beyond the scope of this paper
and can already be found in Faini et al. (1989). In what fo.ows we
only present a stylized description of the main features of the model.
The latter is based on a simple variant of the aggregate supply-aggregate
demand open economy framework. Noticeable features of the model
include the emphasis on supply behaviour, the modelling of import
demand under rationing and the analysis of both external and domestic
debt dynamics.   The presentation is organised by economic agents
(firms, households, goverment, etc.) rather than by the more usual
approach based on a distinction between markets (goods, labour, and
money markets). This should permit a more critical evaluation oF the
microeconomic foundations of the model.
a) a macroeconomic model
Table 6 can used as a guide to the maln relationships comprised In
the model. The first block of equations focusses on the firm. The
approach used to specify the behaviour of the representatlve firms was
described in the previous paragraph when analysing the investment
decision. In eq. 1 in the Table investment (I) is a Function of expected
output (Ye), the cost oF capital (c) and the availability of credit (B ).
The cost of capital (eq. 2) is a function of tax (,u and r) and depreciatfon
(8) parameters as well as oF the ILnterest rate r and the price of
investment goods, p. In turn p (eq. 3) is equal to a weighted average
of the price of domestic (pI ) and imported capital goods, with the
latter depending on the real exchange rate (X, defined as the ratio of
foreign to domestic prices) and on the tariff rate on domestic capital
goods (r .). The specification of expected output is taken from eq. 11 in
the textm'After choosing total investment, the firm decides how to
allocate it between domestic and foreign capital goods as a Function of
their relative prices and of the extent of quantitative restrictions on



26
imForts of capital goods. q1 (eq. 51. Short-run choices can be be
described as fcIlows. Output Isupplv (Yi depends on total capital stock
and the real e<-hange rate, where the latter acts as a proxy for variable
costs. A real depreciation, i.e. an increase in A, will boost wage and
intermediate inputs costs and lowers supply. Given that capital stock
data are not a,ailable In Morocco, we take a first quasi difference of
the origina'l supply funxtion and estimate equation 6. In the estimation
we take the primary and the government sector output levels to be
exogenous. After output supply is determined, the demand for
intermediate inputs (M ) and the supply of exports (X) can be simply
described in eqs. 7 andn8 respectively as a function of Y, the wage rate
(w) and the relevant price variable (p   and p ). We do not however
assume the country to be small in expoRnmarketX. The price of exports
p  depends therefore on exports volumes, world demand (WD) and
foreign competitors' prices (p*). This completes the description of the
firm's behaviour.
Households' choices are described in the second block of equations.
We make a separability assumption where consumers first determine
the level oF saving and aggregate consumption and then allocate total
consumption between domestic and foreign commodities.  Aggregate
private consumption is modelled as in the previous section (eq. 10). In
the present version of the model we take the seciFication where private
consumption depends on disposable income (Y), real interest rate (r)
and the real exchange rate ()\). The allocation oF aggregate consumption
between domestically and Foreign produced commodities is described by
a simple (constrained) linear expenditure system. We allow for the fact
that a subset of foreign consumption goods cannot be freely imported
and study the impact of these rationing measures on total Import of
consumption goods. Our approach parallels the one used for investment
imports and draws on Bertola and Faini (1990). As with investment
goods, we do not allow for any rationing effect on aggregate
consumption.  The resulting equation is highly non-linear, but can be
simply described (eq. 11) as stating that imports of consumption goods
depend on total consumption, relative prices and the extent oF quotas on
such import category (q_). Eq. 12 defines disposable income as the sum
of GCDP, net factor irncome from abroad (NFI), interest on public
domestic (GOINT) and foreign (OXINT) debt, transfer (TR) minus total
taxes (Ti and monetary financing (MF). Note that interest payments on
foreign debt (i XINT) are already included, with a negative sign, in NFI.



27
Then NFI+±CX1N  is a measure cf net factrr payments from  abroad
accruing to the private sectc*r (mainly ',vorkers remittancesi.  Finally
eq. 13 deFines private saving.
The behaviour of the gc vernment sector is described very simply by a
set of accounting and technical identities. As far as taxation Is
concerned, we distinguish between trade (eq. 14) and income (eq. 16)
taxes.  Import duties are estimated endogenously in the model by
applying the relevant duty rate to imports of consumption, investment,
intermediate and other commodities. We also allow for the temporary
levy on petroleum (eq. 15), which was introduced in 1986 and should be
gradually phased out over the next few years. Information on the
revenuF likely to be generated by this levy was provided by IMF
source-. Finally the income tax rate is determined r esidually from
total revenue data. Eq. 17 defines total tax revenue. Government
expenditures are also fairly disaggregated. We distinguish between
government investment (eq. 19), government current spending on goods
and services (eq. 20), current transfers (eq. 18) and interest payments
on both domestic (B  in eq. 21) and foreign (XD in eq. 22) debt. We
can therefore acccdnt in our simulatlons for debt dynamics. The
accumulation of foreign debt is determined by the current account
deficit and by real exchange rate variations which induce a capital gain
or loss in the value of outstanding external debt (eq. 25). The change In
domestic debt (A B ) is equal to that part of the budget deficit which
cannot be financed aroad (eq. 26). Simply manipulations show that A
B  can be also expressed as the difference between private saving and
pgivate investment. The last item  in the consolidated government
Central Bank budget identity is revenue from seignorage and the inflation
tax (eq. 24).
Government revenue from seignorage and the inflation tax is
determined by the private sector's choice of assets. Following the
analysis in section 2, the demand for currency (eq. 27), for demand
deposits (eq. 28) and for time deposits (eq. 29) is simply modelled as
a tf,nction of income, inflation (7r) and the interest rate on time deposits
(i ). In eq. 30 monetary financing (MF) is defined as the change of
narrowly defined monetary base, i.e. of currency and of demRpd
deposits the latter multiplied by the required reserve ratio (rr ).
Finally conditions in the money market will help determine, through the
banks balance sheets identity, the equilibrium in the credit market. The



28
sUm of demand and time deposits define total commerclal banks
liabilities, which are allocated on the assets side among compulsory
reserves (R), credit to the private sector (B ) and credit to the
government (B ) (eq.  3 1  It is assumed that bnks do not hold free
reserves. Crecgt to the government has already been determined by the
governement budget constraint for a given current account deficit and an
inflationary target. From eq. (31) we find that credit to the prlvate
sector is determined residually.  It will In turn affect, as was
mentioned earlier, the demand for investment. Therefore, as clalmed,
fiscal policy will affect Investment not only through Its impact on
interest rates, but also, more directly, by influenclng the availabillty of
credit.
To close the model we need to specify the equllibrium condition In
the goods markets (eq. 33) where total demand, I.e. the sum of
consumption, Investment, government spending on goods and services
and the resource balance (i.e. total exports X mlnus total Imports M) 3,
must be equal to aggregate supply (Y). A second condition relates to
the current account constraint (eq. 34) which is Identlcally defined as
the sum of the resource balance and net factor Income from abroad
(NFI). It can be easily checked that:
Sp + Sg - CA = I + Ipub                                   (13)
i.e. the sum of private, government and foreign savlng (the negative of
CA) is, in equilibrium, equal to total investment.
This completes the description of the model. In the next section we
shall discuss some stylised properties of the model whlch will then be
used to perform some simple comparative static exercices on the Impact
of fiscal policy.
b) simulating the impact of fiscal policy
Table 7 presents the complete model In summary form. There are 34
equations and 35 endogenous variables (33 left-hand slde variables,
allowing for the Fact that Y appears twice, plus the real exchange rate
and the real interest rate). In this sectlon we present a number of
simulations on the impact of fiscal pollcy. The common assumption
underlying all the simulations Is that the economy Is rationed on



29
international markets. As a result the financeable current account
deficit is given at each point oF time, I.e. CA Is exogenous. Notlce
that also the inflation rate is exogenously determined, at a level set
by the Central Bank. Control on the money supply allows the Central
Bank to achieve its inflatlon target.
As in the model by van Wljnbergen (1989), two equllibrium
conditions play a cruclal role In determining the equlllbrium In the
economy. One is the goods market equillbrlum (eq. 33 In the Table),
which can be read as requlrlng that demand and supply for domestically
produced goods be equal. The second condltlon Is the current account
constraint, where the excess of prlvate and publlc saving over total
investment must be equal to the exogenously glven level of the current
account. This condition can be derived by comblnlng eqs. 12- 13, 25-26
and 33-34. The real exchange rate and the real Interest rate wlll move
to equllibrate the goods markets and ensure that the current account
constraint is not violated.
Some further observations on the role of credit constraints and the
interest rate are in order. It was shown earlier how both these factors
will afFect investment. Suppose that the credit constraint Is blnding,
i.e. B*<y in the notation of sectlon 2. Suppose also, but only for a
moment, that investment, and banks llabilltles, are given.  Then an
increase in the interest rate would not move the credit market closer to
equilibrium. As shown In section 2, the composition of Investment
finance (as desired by the flrm) does not depend on the Interest rate but
only on the risk premium. However, in the context of our model, the
increase in r will reduce Investment, until all of it can be financed
from available sources. The firm, though, Is stlll off Its notlonal
demand for credit, being unable to achieve the deslred composltlon of Its
investment finance. Desplte the exlstence of credit constraints, interest
rate variations still play a role In bringlng into equilibrlum saving and
investment.
We pursue three simple simulation exerclces. Column 1 In Table 7
presents the base solution. In the first comparative static exercice
(column 2) we study the impact of an increase In government spending by
MDH 3000. Foreign financing and domestic Inflation are assumed to be
unchanged.  From  eq. 26 in Table 6, we then see that most of the
financing will rely on domestic debt. The increase in goverment spending



30
v.11i bcth cause an escess demand for domestic goods and lead to a
derline in sa\wing and thus a current account imbalance. As expected, In
the new equilibrium (cclumn 2 in Table 7), the real exchange rate will
appreciate (because of the excess demand for domestic goods) and the
real interest rate will increase (because of the fall in saving). Growth
increases somewhat, but the impact of the expansionary fiscal policy Is
mostly reflected in a crowding out of Investment. Capital accumulation
falls partlv because of higher interest rates (whose effect on the cost of
capital is however mitigated by the real appreciation) and mainly
because oF lower availability of credit.
Suppose now that the increase in government expenditure Is financed
from abroad. We model this by assuming that the current account is
allowed to deteriorate by an amount equal to the increase in publlc
consumption. The outcome oF this simulation is presented in column 3.
The higher current account deficit is equivalent to an increase in foreign
saving. We would then expect the interest rate to Increase less than in
the case where no extra financing from abroad was available. This Is not
however what we Find in the third column of Table 7. The real interest
rate is now equal to 3.7% against 2.9% in the previous simulation. The
reason for this somewhat surprising result hinges on the crucial role the
credit market plays in this model. What is ha.pening is that the
possibility for the goverment to finance a larger pcrtlon oF its deficit
abroad reduces the pressure that government borrowing p 's on credit
markets, leaving a larger share of total domestic credit to the private
sector. Investment therefore is not crowded out through credit rationing
and declines less than in the previous simulation. A larger increase In
the interest rate as well as a more sustained real appreciation are
required to bring equilibrium  in the goods market.  When, as in the
First simulation, the government higher expenditure was financed
domestically, most of the crowding out of investment took place through
credit rationing with more limited effects on the interest rate. The
difFerences in outcome between the two simulations highlight the crucial
rcle that conditions in the credit market play in determining the impact
of fiscal policy.
Finallv in the last column oF Table 7 we consider the case where a
foreign financed increase in government expenditure Is accompanied by
a process of financial liberalization. We model this process as a
reduction in the liquidity requirement imposed on demand deposits.



31
TABLE 7
The complete modell
a) the firms
(1)  1 = j(Ye c B )                      private investment demand
(2)  c = p1 (1-,p) [r (1-r) + 6                       cost of capital
(3) PI  Pd Pdi + (1-$d) (1+tmi) X           invest. goods price
(4)  ye  Ye(w(_1), c(-1), Ipub(-1), MS(-1))         expected output+
(5)  MI= M 1(1 Pdi/Pmi' q1)             imports of investment goods
(6)  Y = Y( Y(-1), 1, X, X(-1))                      supply of goods
(7)  Mn = Nn(Y, W/Pmn)                imports of intermediate goods
(8)  X = Xs(Y, px/w)                  export supply of manuf. goods
(9)  px  p P(X, WD, p*)               export demand of manuf. goods
b) the households
(10)  C = C(Yd, r, X)                          private consumption
(11)  Mc =  C(C, Pdc/Pmc qc)           imports of consumption goods
(12)  yd - Y+NFI+GDINT+GXINT+TR-T-MF             disposable income
(13)  S  =yd  c                                      private saving
p
c) the government
(14)  TM. = r. k M.                       import taxes (i=n,c,o,i)
(15)  TP = tp Y                                     petroleum tax
(16)  TY = t Y                                        other taxes



32
(17)  T = z    TM  + TP + TY                           total taxes
(18)  TR = t Y                                          transfers
(19)  I= 1p     Yb Y                             public Investment
(20)  0 = g Y                govt. expendlture on goods and services
(21)  ODINT = idd Bg(-1)            Interest payments on dom. debt
(22)  GXINT = ixd XD(-1)          Interest payments on forelgn debt
(23)  A Idd = A r + A Tr            Interest rate on domestlc debt
(24)  S = T-GDINT-OXINT-TR-G+MF                government savlng
(25)  XD = XD(-1) X / X(-1) - CA                 external debt
(26) A Bg - CA = Ipub - Sg                 govt. budget constraint
c) the credit and the money markets
(27)  CUR = CUR(Y, Itd, Tr)                 demand for currency
(28)  DD = DD(Y, itd , 7r)            demand for demand deposlts
(29)  TD = TD(Y, ltd, ir)                demand for time deposits
(30)  MF  A CUR + rrdd A DD                   monetary financlng
(31)  Bp  DD+TD-(R+B )                credit to the private sector
(32)  A itd = A r + A Tr           Interest rate on tlme deposlts
d) natlonal income fdentitles
(33)  Y = C+G+X-M+I+I pub-- TM        goods market equilibrlum
(34)  CA = X - M + NFI                          current account



33
Notes:
+ not implemented In the present version of the model
Other symbols:
\: real exchange rate
r: real interest rate
MS: money supply
q.: extent oF QR's on Imports of type I
EB2: credits to firms
? mj foreign currer.cy price oF Imports of type j
Pmj domestic currency prlce of Imports of type j
Pdj domestic price of good j
px: export price
r  tariff rate on good j
r: corporate tax rate
w: wage rate
WD: world demand
NFI: net Factor Income from abroad
7r: inflation rate
R:dbanks reserves
rr = reserve requirement for demand deposlts



34
There are certalnly beneficial influences from a more competitive and
less heavily regulated financial sector, which are not however modelJed
in our set-up. But there are also some macroeconomic costs as shown In
column 4 of Table 7. The direct Impact of the reductlon In the liquldity
requirement Is to Increase Interest expenses on public domestic debt and
force the government to rely to a larger extent on domestic credit
markets. The consequent decline In Investment relleves the pressure on
the domestic goods markets (Indeed the real exchange rate appreciates
by less than In the second simulatlon), but does not mitigate the Impact
r-n Interest rate of hlgher public spending. Agalnst the long-run benefits
n financial liberalization, some allowance should therefore also be
ie for the macroeconomic costs stemming from a further
.erioratlon In the situation of the government budget.



35
TABLE 7
Slmulatlng the Impact of fiscal policy
base run   domest. financed  foreign financed  financial
increase in G    increase in C    liberaliz.
(1)           (2)              (3)          (4)
r            2.7           3.1              4.3           4.4
X           100            96.7             87.5         86.5
i/Y         18.4           16.1            16.5          15.8
BD/Y        -4.8           -6.8            -7.3          -8.1
l"          6.7            7.1             8.3            8.4
B p/Y       21.4           19.2            19.4          18.6
GDINT/Y      2.5            2.6             2.7            3.4
Legenda: r: real interest rate, X: real exchange rate, I/Y: GDP share
of prlvate Investment, BD/Y: GDP share of budget deficit; Y': growth
rate, B /Y: GDP share of credit to the prlvate sector, GDINT/Y:
GDP sh&re of goverment Interest expenses on publlc domestic debt.
Slmulatlons:
1: base case.
2: Increase ln government current consumptlon of MDH 3000.
3: as in 2, plus a current account deterloratlon of MDH 3000.
4: as In 3, plus abolition of the llquldlty requlrement on demand
deposits.



36
6. CONCLUSIONS
Morocco has made great pro ress toward macroeconomic and fiscal
stabllity. Yet the need to consolidate and broaden the achievements to
date remains paramount. We have argued In this paper that even the
financlng of a relatlvely unchanged tudget deficit may pose major
problems. It is unlikely that foreign finance wlll increase substantially
in the medlum-run. Monetary financlng does not seem to be a palatable
alternative given its highly inflatlonary impllcations (sectlon 3). Flnally
lncreaslng rellance on domestic financlal markets ls likely to lead to a
steep increase In Interest costs for the Treasury wlth a destabillsing
effect on the evolutlon of the main publlc debt Indlcators.  Our
slmulatlons also suggest that an Increase In government current
expenditure crowds out Investment. The short-run benefits on output of
such a pollcy may then be outweighed by Its long-run negative Impact on
growth.  It ls essential therefore that the commltment toward flscal
disclpllne remalns unshaken. At the same tlme a determlned effort Is
requlred to Implement an effectlve reform In the tax and public
expendlture system, wlth a vlew of avoldlng that the brunt of flscal
adjustment falls agaln mostly on public Investment.
At a more general level, we have argued that the effects of budget
deflclts cannot be measured by looklng only at the Impact they exert
through changes in the Interest rate on aggregate demand. As shown by
Blinder (1987), Bernanke and 2linder (1988) and Blinder and Stlglitz
(1983), a large part of the Impact of flscal pollcy may work Its way
through the credit market. An expanslonary flscal pollcy may
exac_rbate the pervasiveness of credlt ratlonlng effects on Investment
demand wlth limited Impact on the level of the Interest rate. Our
estlmates support the clalm that, even after controlling for the cost of
capital, the avallabllty of credlt plays a slgnlflcant role In Influenclng
the demand for Investment. As suggested by our slmulations, the credlt
channel In turn will be a cruclal factor In determilnlng the
macroeconomlc outcome of different flscal policles. Flnally we argue
that the Impact on macroeconomlc equlllbria, In partlcular on the
government budget, should be a relevant factor In assesslng the speed of
financlal liberallzatlon.



37
ENDNDT ES
I F^r a seminal contribution to the emplrical evaluation of the
relationship between capital market Imperfections and investment
decisions, see Fazzari et al. (1988).
2 The formal solution tc the firm's optimization problem is available
from the author.
3 In the empirical implementation of the model only manufacturing
expcrts on the one hand and investment, intermediate and consumption
goods imports on the other are endogenous. All the other components of
the resource balance (mainly exports of phosphate rocks and
agricultural goods, imports of food and petroleum, net tourism
receipts) are assumed, in the present version of the model, not to
depend on price incentives and are, as a result, projected exogenously.
Some of these components, in particular tourism receipts, may be
endogenised in a new version of the model. Simllar remarks apply to
the flow of workers' remittances (a component of NFI) which should not
be treated as exogenous, but should be related to the level of the
interest rate and the (expected) movements of the exchange rate.



38
REFERENCES
Bernanke B. and A. Blinder (1988), "Credit, money and aggregate
demand", NBER Working Paper n. 2534.
Bertola C. and R. Faini (1990) "Import demand and non-tarlff barrlers.
The impact of trade liberalization", Journal of Development Economlcs,
forthcoming.
Blinder A. (1987), 'Credit rationing and effective supply fallures",
Economic Journal, 327-352.
Blinder A. and J. Stiglitz (1983), "Money, credlt constralnts and
economic activity", American Economic Review, 297-302.
Deaton A. (1989). "Saving in developing countries: theory and revlew",
Conference on Development Economics, Supplement to the World Bank
Economic Review, 61-96
Faini R. and J. de Melo (1990), "Adjustment, investment and the real
exchange rate", Economic Policy, 11, 491-519.
Faini R., J. Porter and S. van Wijnbergen (1989), "Trade liberalization,
budget deficits and growth", mimeo.
Fazzari  F.M., 0. Hubbard  and  C. Petersen  (1988), "Financlng
constraints and corporate investment", Brookings Papers on Economic
Activitv, 19, 141-195.
Haque N. (1989), "Fiscal policy and private saving behavlour In
developing countries", IMF Staff Papers , 316-335.
Haque N. and Montiel P. (1989), "Consumption in developlng countrles:
tests for liquidity cor:straints and finIte horizons", Revlew of
Economics and Statistics, 408-415.
Hendry E). and T. von Ungern Sternberg (1981), "Liquldlty and Inflatlon
effects in consumers expenditures", in A. Deaton (ed.), Essays In theory
and measurement of consumer behaviour, Cambridge Unlverslty Press.



39
Nam S. (1989), "What determines national saving ? A case study of
Korea and the Philippines', The World Bank, PPR WP 205.
Pagan A. (198'4), "Econometric ssues in the analyis of regressions with
generated regressors", International Economic Review. 25, 1, 221-247.
Rossi N. (1988), "Government spending, the real interest rate and the
behaviour of liquidity-constrained consumers In developi g countries",
IMF Staff Papers, 35, 104-140.
Schmidt-Hebbel K. and T. Muller (1990), "Private investment and
private saving in Morocco", The World Bank, mimeo.
United Nations-World Bank (1990), Morocco 2000. An open and
competitive economy, Trade Expansion Program.
van Wijnbergen S. (1989), "Growth, external debt and the real exchange
rate in Mexico", The World Bank, PPR WP 257.
World Bank (1990), Morocco: Sustained investment and growth in the
nineties, Emena, Country Operations Division.



PRL WorKino Paoer Series
Contact
1:                       ~~~~~~~~                           ~~Author  _for p2aper
WPS6o7  Abol!l,h-iq ('1'1!" Ritue   The EttectJ Donald F. Larson      March 1991        D Gustafson
on Cereals. Sugii and Oilseoed,   Simon Glance                                33714
in West Germnw-y                  Brent Borrell
Merlinda Ingco
Jonathan Coleman
WPSGo8  Cross Count,y Studies of Growth     Ross Levine               March 1991        CECMG
and Policy. Me!hodo!og;cal,       David Renelt                                39175
Conceptuai, ind Statistical Prob,ems
WPS609  A Sensitv'ty A'ralysis of Cross-    Ross Levine               March 1991        CECMG
Country Growth Regressions        David Renelt                                39175
WPS610  Can Preshiprnent Inspection Offset  Alexander J. Yeats        March 1991        J. Jacobson
Noncompetitive Pric ng of Developing                                          33710
Countries Imports9 The Evidence
from Madagascar
WPS611  Tariff-based Commodity Price        Jonathan R. Coleman       March 1991        S. Lipscomb
Stabilization Schemes in Venezuela  Donald F Larson                           33718
WPS612  Eduration and Productivity in       Lawrence J. Lau           March 1991        WDR Office
Developing Countries: An Aggregate  Dean T. Jamison                           31393
Production Function Approach      Frederic F. Louat
WPS613  Price-Wage Dynamics and the          Simon Commander          March 1991        0. Del Cid
Transmiss!on of Inflation ir, Socialist  Fabrizio Coricelli                   39050
Economies. Empirical Modelb or
Hungary and Poland
WPS614  Accountability in Public Services:    Samuel Paul             March 1991        E. Madrona
Exit, Voice. and Capture                                                      37496
WPS615  Socialist Economnic Growth and       Heng-fu Zou              March 1991        A. Bhalla
Political Investment Cycles                                                   37699
WPS616  Optimal Nonlinear Income Taxation   Ravi Kanbur               March 1991        J. Sweeney
for the Alleviation of Poverty     Michael Keen                               31021
Malt; Tuomala
WPS617  International Poverty Protections   Sudhir Anand              March 1991        J. Sweeney
Ravi Kanbur                                31021
WPS618  Poverty and Development The         Ravi Kanbur               March 1991        J. Sweeney
Human Development Report and the                                              31021
World Development Report. 1990
WPS619  Foreign Direct Investment in         Laurence Cockcroft       March 1991        S. King-Watson
Sub-Saharan Africa                Roger C. Riddell                            33730



lI Working Paper Series
Coontact
liLlffl Author                                            lof papei
WPS620  Have Commercial Banks Ignored      Sule Ozler               March 1991        S King-Watson
History?                                                                    33730
WPS621  Sensible Debt Buybacks for Highly   Enrica Detragiache      March 1991        S King-Watson
Indebted Countries                                                          33730
WPS622  How Factors in Creditor Countries    Sule Ozler             March 1991        S King-Watson
,.ffect Secondary Market Prices  Harry Huizinga                             33730
for Developing Country Debt
WPS623  World Bank-Supported Adjustment   Vitt3rio Corbo            March 1991        A. Oropesa
Programs: Country Performance     Patricio Rolas                            39075
and Effectiveness
WPS624  Choosing Policy Instruments for    Gunnar S. Eskeland       March 1991        A Bhalla
Poliution Control: A Review      Emmanuel Jimenez                           37699
WPS625  How Trade and Macroeconomic        Rarnon Lopez             March 1991        K. Cabana
Policies Affect Economic Growth                                             37946
and Capital Accumulation in
Developing Countries
WPS626  The Macroeconomics of the Public   William Easterly         March 1991        R. Luz
Sector Deficit: The Case of                                                 39059
Colombia
WPS627  The Role of Institutions in Poverty    Sharon L. Holt       March 1991        WDR Office
Reduction: A Focus on the                                                   31393
Productive Sectors
WPS628  The Indonesian Family Planning     Dov Chernichovsky        March 1991        0. Nadora
Program: An Economic Perspective  Henry Pardoko                             31091
David De Leeuw
Pudjo Rahardjo
Charles Lerman
WPS629  An Atheoretic Evaluation of Success  Patrick Conway         March 1991        D. Ballantyne
in Structural Adjustment                                                    37947
WPS630  The Spirit of Capitalism and Long-run  Heng-fu Zou          March 1991        A. Bhalla
Growth                                                                      37699
WPS631  The Macroeconomics of the Public    Riccardo Faini          March 1991        R. Luz
Sector Deficit: The Case of                                                 34303
Morocc o
WPS632  The Macroeconomics of the Public   Carlos Altredo Rodriguez   March 1991      R. Luz
Sector Deficit: The Case of                                                 34303
Argentina