Untitled-10

Bangko Sentral Review July 2004

EFFICIENCY OF FISCAL AND MONETARY EFFICIENCY OF FISCAL AND MONETARY

EFFICIENCY OF FISCAL AND MONETARY

POLICIES IN THE PHILIPPINES:POLICIES IN THE PHILIPPINES:

POLICIES IN THE PHILIPPINES:

THE ST. LOUIS MODEL APPROACH THE ST. LOUIS MODEL APPROACH

THE ST. LOUIS MODEL APPROACH

Neil Angelo C. Halcon

Leah Melissa T. De Leon

Introduction

ebates have persisted around what policy is

deemed effective in stimulating as well as in

stabilizing growth in the Philippines. Over time, the

interactions of both fiscal and monetary policies have

largely depended on the structural adjustments and

the reformation of government and financial

institutions. However, fiscal policies in the past were

influenced by structural deficits and internal economic

volatilities caused by political upheavals in the last

two decades. Similarly, the vulnerabilities of large

monetary shocks bolster the impression that the

country’s monetary strategy is less than optimal. While

a policy mix feasible enough to achieve the right

balance of macroeconomic stability and growth

remains an elusive dream, it is hoped that the policy

implications arising from the study and relevant to

the Philippines would encourage better management

of these major policy instruments.

It would be useful, therefore, to undertake an

empirical investigation into the efficiency of both fiscal

and monetary policies at achieving the twin macro

objectives under the internal equilibrium of stable

growth and low inflation within the framework of the

St. Louis Model. This paper also reviews briefly the

issues around which the fiscal and monetary policies

are hinged, as well as highlights the fiscal and

monetary policy scenario of the Philippines.

Issues on Monetary Policy Efficiency

Although monetarism is as alive and well as ever,

considerable skepticism and contrary opinions can

be found that surprisingly echoes the vital flaw in

human reasoning: expectations. The twin realization

that it is impossible to observe expectations directly,

and that reasoning is supposed to be based largely on

experience, made monetarists confront the issues of

credibility and sustainability among decision-makers.

Expectations. The monetarist theoretical

scenario has a strong policy message. The existence

of an “expectations component” in the argument

means that lags of various kinds exist within the

implementation of monetary policy. The central bank

in itself has both the inside and outside lags to contend

with. Inside lags exist because of administrative

delays and delays in the recognition of adverse

macroeconomic developments in output, employment

and prices. While these may be of shorter actual

duration than in the case of fiscal policy, the outside

lag – the length of time it takes before the actual

changes in monetary expansion/contraction are felt

on the “target” variables of inflation, output, and

employment – is very significant.

In the monetarist view, expectations adjustment

is a time-consuming process. Comparatively little is

known, however, about the formation of expectations

and other factors affecting the length of lags. For

that reason, a good deal of uncertainty accompanies

the conduct and effectiveness of monetary policy.

Dynamic inconsistency and inflationary bias.

This refers to the difference between the optimal

policies that a central bank would announce if it were

considered credible by the public, and the policies it

would carry out after the public had made decisions

on the basis of its expectations. In reality, however,

the public can discount the announcements of the

central bank, and the resulting inflation rate will be

higher than it needs to be. As a result, output may or

may not rise above the full employment rate,

depending on current wage rigidities in the system

that prevent complete wage and price adjustments.

The incentive of policymakers to promote low inflation

July 2004 Bangko Sentral Review

is constrained by the behavior of rational agents,

creating an economy with inflationary bias.

Rules versus discretion. The old debate on

rules versus discretion is related to the behavior of

central banks in the conduct of monetary policy: rule-

like behavior implies a systematic conduct of policy

without exploiting the existing expectations to achieve

temporary gains in output. Discretion, on the other

hand, considers the choices of different periods (time-

consistent strategy) and offers mainly a shortsighted

solution. As time-consistent policy may bring about

significant short-run social benefits, economic agents

will learn to anticipate the period-by-period

optimization, vitiating the credibility of the policymaker.

Rules would have provided for more useful

information about the stance of monetary policy during

particular episodes and a credible central bank is

essential to achieve this.

Sustainability of monetary policy in uncertain

times. In a speech made by William Poole of the

Federal Reserve Bank of St. Louis, he emphasized

the inflation objective of the Fed, which is to maintain

a low and stable rate of inflation. The reason for

emphasizing this goal is that the U.S. economy’s long-

run economic performance in terms of employment

growth and economic growth is maximized when the

rate of inflation is low and stable. Moreover, no other

economic policy authority can achieve the inflation

outcome. Hence, controlling the creation of money is

its main responsibility, and exercising that power wisely

is its main monetary policy function.

Interestingly, Poole acknowledged the fact that

economic historians studying monetary policy in the

United States and other countries have argued that

central banks have from time to time made serious

mistakes that increased rather than reduced

fluctuations in the unemployment rate. Certainly, the

first obligation of a central bank is to do no harm.

Over time, advances in economists’ understanding

of macroeconomics and monetary policy have led to

improvements in monetary policy.

These observations make clear that the issue of

the length of the lag from the Fed’s policy action is

not well defined, because the market gradually eases

its rates in anticipation of eventual Fed action. What

central banks can do is to respond sensibly to current

developments, making sure that long-term policy goals

are kept in place and there are no short-run

disturbances nor overreactions that will have

unfavorable effect on the economy.

Issues on Fiscal Policy Efficiency

The macroeconomic relationship between fiscal

policy and economic growth has long fascinated

economists. Indeed, fiscal policy has been the major

policy instrument for macroeconomic management.

However, it has been found to be a clumsy instrument

with its attendant delays in the start-up of investment

projects, which includes the preparation and floatation

of tender documents and approval of contracts as

well as uncertain impacts due to lags in project

implementation.

Role of fiscal policy in recession. Recent

years have seen a revival of the debate about the

role of fiscal policy in stimulating economic activity,

particularly given the recessions in Asian crisis

countries, the prolonged slump in Japan and, more

recently, the slowdown in the United States (Lane, et

al., 1999). While fiscal policy in the Asian crisis

countries became increasingly oriented toward

supporting economic activity, questions still remain

about the effectiveness of fiscal stimulus during a

crisis. Even if it is generally agreed that there are

circumstances where fiscal policy cannot be loosened

(e.g., when fiscal imbalances or debt sustainability

problems are the root causes of the crisis), whether

and when expansionary fiscal policy is effective in

supporting activity need to be studied further.

Overall, many of the observations appear to be

broadly in line with theoretical predictions and the

following stylized facts emerge (Baldacci, et al.,

2001):

• On average, fiscal policy is expansionary during

recession episodes. However, a very large

number of recession episodes (40 percent of

total) are accompanied by contractionary fiscal

policy. Unfavorable initial conditions (high

public debt, and large fiscal and current account

deficits) are associated with a contractionary

fiscal response in a recession, while negative

terms of trade shocks or a large public sector

tend to result in a more expansionary fiscal

response.

Bangko Sentral Review July 2004

• There is some evidence that expansionary fiscal

policy dampens the severity of a recession,

especially in open economies with a fixed

exchange rate, favorable initial fiscal and

external conditions (low public debt, a large

public sector, and a small current account

deficit), and in combination with expansionary

monetary policy. Expenditure-led fiscal

expansions are also associated with less severe

recessions. All these results are consistent with

the predictions of the theory, even though the

magnitude of the average effects is small, and

the variance is large.

• There are marked differences between

advanced economies and other country groups.

The fiscal response is more often expansionary

in advanced economies. At the same time, the

impact of fiscal policy tends to be smaller than

in other groups.

• A number of other factors are associated with

both the fiscal response and the severity of a

recession. Inflation tends to be higher, and

monetary policy more contractionary during

episodes with a contractionary fiscal response,

which makes the interpretation of the results

somewhat difficult and points to the need for a

multivariate approach.

Apart from the empirical studies confirming the

superiority of fiscal over monetary policy efficiency,

equally important also is the role of the central bank

in conducting monetary policy. The latter has attracted

arguments as to the ineffectiveness of the policy and

in assessing the general framework of monetary

policy, (Barth, 2002) discussed current issues

highlighting the fallacies inherent in the framework:

dynamic inconsistencies resulting from rational

expectations, and the long standing issue on rules

versus discretion.

Crowding out effect. There have been three

distinct sets of arguments to the effect that fiscal

policy will be ineffective, under the general heading

of “crowding out”. The first, in the context of the IS-

LM analysis, was a “crowding out” due to a rise in

interest rates following a fiscal expansion. This was

based on an exogenous money supply and the interest

rate equating the demand for and supply of money.

The second form of “crowding out” arose from a

combination of the notion of a supply-side equilibrium

(i.e., NAIRU) and that the level of aggregate demand

would adjust to be consistent with the supply-side

equilibrium. As has been argued above, fiscal policy

has an effect on the level of aggregate demand, and

“crowding out” only occurs if it assumed that the

supply-side equilibrium must be attained and that level

of aggregate demand would be equivalent to the

supply-side equilibrium. The third route comes from

the “Ricardian equivalence” proposition. An expansion

of government expenditure, however funded, is

postulated to lead to an equivalent reduction in private

expenditure, leaving the overall level of demand

unchanged (Arestis and Sawyer, 2003).

Fragile fiscal policy indicators. Unfortunately,

analyses of fiscal policy indicators have frustrated

empiricists for almost as long (Fu, et al., 2003). One

root of that frustration is the array of possible policy

indicators. As earlier discussed in Tanzi and Zee

(1997), there are three candidate indicators of fiscal

policy—government expenditures, taxes and deficits.

The literature does not systematically favor one

indicator of fiscal policy over the others. Furthermore,

Levine and Renelt (1992) find that none of these fiscal

indicators is robustly correlated with economic growth

when evaluated individually.

The fragility of fiscal indicators found in Levine

and Renelt (and contradictory findings in the growth

literature in general) probably arises from the inability

of any single budgetary component to fully capture

the stance of fiscal policy. For example, an increase

in government expenditures could be considered

expansionary if it were financed by deficit spending.

However, it could also be considered contractionary

if it were financed by an increase in taxes because

such a policy would imply an increase in the size of

the public sector.

Keynesian versus Monetarist. The Keynesian

tradition gave government the responsibility of

stabilizing a troubled economy. Keynesians developed

the notion of a fiscal/monetary mix to control spending

and the balance of payments simultaneously. Judicious,

well-timed changes in taxes and government spending

were to be balanced against propitious changes in

money to manage the economy. The famous Phillips

curve trade-off supposedly gave economists a tool

July 2004 Bangko Sentral Review

for choosing between inflation and unemployment. If

the choice did not work out as intended, Keynesians

relied on informal price and wage controls, jawboning

(threats), and guideposts to improve the trade-off.

To know when and how much to adjust policies,

Keynesian economists developed forecasting models

that could simulate possible policy changes to predict

their effect and more closely adjust the mix of policy

actions. The Monetarists have always been critical

of these models and their use in policy. They favor

stable policy rules that reduce variability and

uncertainty for private decision makers. They argue

that government serves the economy best by

enhancing stability and acting predictably, not by trying

to engineer carefully timed changes in policy actions.

For the Monetarists, such efforts are frequently

destabilizing (that is, doing the opposite of what they

were supposed to do). Thus, the attempt to apply

Keynesian policies, notably in the United States and

Britain produced alternating periods of rising inflation

and rising unemployment, and not the finely adjusted

trade-off that the Keynesians sought.

No perfect forecast. Forecasting proved a

weak foundation for policy actions. The best forecasts

of spending, output, prices, and inflation proved to be

unreliable. Systematic studies of forecasting accuracy

show that on average forecasters have been unable

to distinguish between booms and recessions a quarter

or a year ahead, so they are as likely to mislead as to

benefit policymakers. The records of the Federal

Reserve that have become available show that during

the period of rising inflation, annual inflation was

always under predicted. When inflation fell in the

eighties, the Federal Reserve persistently predicted

too high an inflation rate. A vast amount of research

has shown that econometric models cannot accurately

forecast interest rates and exchange rates.

Philippine Case: Structures and

Arrangements

Fiscal policy. Fiscal balance is important in

achieving economic stability, and there are two

important issues that must be addressed: (1) the

sustainability of the fiscal balance, and (2) the impact

of government spending and debt on the economic

growth (Paderanga, 2001).

For the sustainability of the fiscal balance, the

ability of the economy to accumulate savings in order

to finance its public investment is very important.

Unfortunately, the slow accumulation of savings by

the public sector has been offsetting the high savings

achieved by the private sector. This resulted in

negative savings-investment gap for 17 consecutive

years. Further, with the growing trade gap, there has

been increasing pressure for the private sector to

generate private savings and likewise for the

government to generate a large primary surplus.

The second issue relating to fiscal policy for

sustainable growth is the impact of government

spending and debt on economic growth. The debt

burden keeps the government from providing services

to the people and the necessary infrastructure to help

improve the economy. The trend in the government’s

budgetary policies is a rising allocation for general

public services and a decreasing budget for economic

services. This means that the government is in essence

spending more for less important matters and spending

less on areas that are more crucial in achieving

sustainable growth.

Fiscal stability after the financial crisis in mid-

1990s remains in jeopardy. To raise its sustainable

growth rate, the Philippines needs to close its fiscal

gap without necessarily sacrificing essential physical

and social infrastructure, i.e., physical and human

capital investment. Posting a well-behaved fiscal

position after the Asian contagion set in, the

government indulged in deficit-spending in an attempt

to stimulate the economy out of a mild recession. Both

the revenue (tax collection) and expenditure sides

require tremendous restructuring efforts and

commitment.

Monetary policy. Despite evidence showing that

only M1 is co-integrated with interest rates, output

and the exchange rate (Gochoco-Bautista, 1993), the

Central Bank of the Philippines (CBP) used M3 or

total liquidity as its intermediate target, specifically

because the ultimate target is the rate of inflation.

As an intermediate target, the CBP used the base

money (BM) as operating target, and the BM is related

to M3 via the money multiplier, i.e.

M3 = money multiplier x BM

Bangko Sentral Review July 2004

In the mid-1980’s, the CBP conducted monetary

policy through monetary aggregate targeting. Under

this approach, its policy actions were aimed at

influencing the behavior of monetary aggregates.

Behind this is the presumption that monetary

aggregates are meaningful indicators of economic

activity, implying that there are stable and predictable

relationships between high-powered money (or the

monetary aggregate that the central bank is able to

control) and money supply, as well as between money

on the one hand, and output and inflation on the other

hand.

However, financial liberalization in 1993 had led

to changes in financial structures and the evolution

of new financial products and services. These

changes have caused large fluctuations in the money

multiplier and the velocity of money such that the

traditional relationships between monetary aggregates

(monetary base and money supply) and between

money supply and the rates of inflation and economic

growth have weakened considerably. Guinigundo

(1999) presented evidence supporting this claim: the

money multiplier increased beginning in 1993 after

financial liberalization; however, there was

deceleration in the rate of inflation from 9 percent in

1994 to 8.1 percent in 1995 despite high rates of

liquidity growth in those years. He attributed this

inflation performance in part to supply-related factors

such as the increase in agricultural output in 1994

and the alleviation of power shortages.

The apparent weakening of these key

relationships under the monetary targeting framework

prompted the Bangko Sentral ng Pilipinas (BSP) to

adopt a “modified targeting framework” beginning

June of 1995, putting greater emphasis on price

stability as well as broadening the information set for

monetary policy decisions. This information set

includes movements in key interest rates, exchange

rate, equity prices, demand-supply indicators and

external economic conditions, among other things.

The BSP’s shift to inflation targeting was

approved by the Monetary Board in 2000 and

implemented in January 2002. Inflation targeting as

monetary policy framework has emphasized the

central role of information in the conduct of monetary

policy. As a forward-looking and an information-

intensive approach to monetary policy, the shift to

inflation targeting would require the use of a wider

set of information variables on the economy than had

been previously utilized. Under inflation targeting,

monetary authorities in the Philippines would have

reliable and timely economic indicators and a good

system for economic forecasting, particularly inflation.

A well-organized system for economic statistics—

based on reliable, timely and comprehensive

information—enable informed decisions and the

efficient functioning of markets. It is also an essential

element in strengthening the monetary policy decision-

making process by providing a focused and consistent

framework for setting the monetary policy stance

(Tetangco and Tuaño-Amador, 2002).

Exchange rate policy. The country’s experience

with fixed and then floating exchange rate regimes

imparts several important lessons. Recall that the lack

of sustainability and implementation of loose financial

policies under the fixed exchange regime resulted in

two large devaluations in the mid-1980’s that brought

about the adoption of a flexible exchange rate in 1985.

Initially viewed to promote external recovery, it

produced mixed results on the inflationary front as

the economy was left without a nominal anchor for

domestic prices. The use of base money targets was

uneven, and the exchange rate was not allowed to

float freely. Apart from imparting an inflationary bias,

the implicit shifts between money supply and

exchange rate targets sent conflicting signals to

markets, affecting the credibility of monetary policy

(Houben, 1997).

The opening of the economy suggested that an

exchange rate anchor could be an effective instrument

to limit inflation. However, a formal exchange rate

commitment would limit monetary policy

independence, making the economy vulnerable once

more to large real shocks. It also depends on whether

the real exchange rate continues to appreciate

(reflecting the rapid structural and technological

improvements).

In such a case, adhering to either a strict

monetary or exchange rate anchor could have high

costs in terms of output growth or inflation. According

to Houben, the Philippine experience points to the

inflation targeting approach as a more viable strategy,

since monetary policy is targeted directly at inflation,

July 2004 Bangko Sentral Review

rather than its intermediate targets. By committing to

more stable prices in a more transparent and consistent

framework, the monetary authority can maintain some

flexibility to cope with unforeseen shocks.

The St. Louis Model

The St. Louis Model is almost an obscure model

to begin with. Developed in the early 1970s, it was

described by Andersen and Carlson as a small-scale

monetarist model of economic activity (King and

Wolman, 1996). It relates the growth of nominal

income (GNP) to changes in monetary and fiscal

actions that are measured by the growth of money

and government expenditures, respectively.

The current version of the St. Louis Model

consists of five estimated equations and a number of

identities. The centerpiece of the St. Louis Model

was the total spending equation, put forward by

Andersen and Jordan (1968), which put together the

change in nominal GNP to changes in the nominal

money stock and to (high employment) government

expenditures. Since the effects of expenditures are

small, the specification embodies the monetarist

viewpoint that monetary change is the key variable

that explains nominal income movements while fiscal

variables only have transitory effect (Meyer and

Varvares, 1981).

Limitations. Although its monetarist background

and structural linkages from money to economic

activity are now widely accepted, the model remains

vulnerable to criticisms due to the essential quantitative

effects of expectations (or Lucas’ Critique) in the

model (Andersen and Carlson, 1970).

Model specification also posted a problem that

concerns specifying the order of the distributed lags

(1, K). Batten and Thornton (1984) furthered that in

any finite distributed lag model, a trade-off occurs

such that a bias is created in specifying too short a

lag (or too low a polynomial degree) against the

inefficiency associated with too long a lag (or too

high a polynomial degree). Either way, the estimates

will be unbiased but inefficient. Since the Andersen-

Jordan equation, and in effect the St. Louis model

employs distributed lag equations, it follows that the

resulting estimates will be biased and may be

inefficient if one lag is too long and the other too

short.

Results of St. Louis model researches. Thus

far, evidence point to the robustness of the St. Louis

equation owing both to the specification of its lag

structure and the imposition of polynomial restrictions.

Taking these into consideration, Batten and Thornton

concluded that in a sample over the period 1962:2 to

1982:3, a one-percentage point increase in money

growth leads to a one-percentage point increase in

the rate of growth of nominal GNP cannot be rejected

at conventional levels of statistical significance.

Alternatively, high employment government spending

has a permanent impact on the rate of growth of

nominal GNP only when contemporary government

spending growth alone is included in the distributed

lag of money growth in the model. In the long run,

monetary policy is effective and fiscal policy is

ineffective in influencing the growth of GNP in the

U.S. economy.

Estimation of the St. Louis equation involving six

developed countries (Canada, France, Germany,

Japan, the UK and the US) was conducted (Batten

and Hafer, 1983). A modified Andersen-Jordan

equation is herein introduced in response to the past

criticism of the original St. Louis equation. Estimating

the modified St. Louis equation for the six countries

yielded results that indicated that changes in money

growth have a significant and lasting impact on

nominal income growth in all six cases. Furthermore,

the money-GNP link was stable in all developed

countries, owing to the then recent move from fixed

to floating exchange rates. On the contrary, fiscal

actions are significant only in the UK and France.

While both studies validated the generalized St.

Louis result, there had been attempts at reassessing

the role of fiscal policy within the framework of the

St. Louis equation (Hafer, 1982). The results are

broadly consistent with previous findings. Specifically,

fiscal actions exert neither a significant nor lasting

impact on the growth of GNP. The results also provide

further evidence against the reliance on fiscal policy

measures to explain movements in GNP. It concludes

by saying that fiscal policy measures (1) do not

significantly increase the explanatory power of an

equation that already incorporates money growth, (2)

Bangko Sentral Review July 2004

do not exhibit stable statistical relationships with GNP

growth, and (3) are not exogenous with respect to

GNP growth.

The Andersen-Jordan Equation

The centerpiece of the St. Louis Model was the

total spending equation put forward by Andersen and

Jordan (1968), that linked the change in nominal GNP

to four-quarter distributed lags of changes in the

nominal money stock and of high-employment

government expenditures. Currently specified in rate-

of-change form, it has the following as independent

variables: nominal income Y, the money supply M

(M1B is the definition of money currently used with

the St. Louis Model), and the high-employment level

of government expenditures G. Dots over the variables

indicate compounded annual rates of change.

The estimates on the M and G variables

approximately sum to unity and zero, respectively.

Hence, these estimates support the general

conclusion associated with a monetarist viewpoint:

Monetary change is the key variable explaining

nominal income movements while fiscal variables

only have a transitory effect.

However, since monetary and fiscal actions

obviously affect the foreign sector, discussion on the

Andersen-Jordan estimates assume the economy

being analyzed is relatively “closed”; that is, its exports

do not account for a large proportion of its GNP. As

such, correlation between external and domestic

influences on GNP is minimized, and external

influences excluded in the analysis (Batten and Hafer,

1983).

As a response both to past criticism of the St.

Louis equation and the likely correlation of domestic

and external influences on GNP, a modified version

of the St. Louis model as conceptualized by Batten

and Hafer (1983) is likewise used.

where: EX = merchandise exports

Y = GNP

M = narrow money

G = government expenditures

The dots above each variable indicate that the

equation is estimated in growth rate form. The

modified St. Louis equation is typically estimated with

each distributed lag’s coefficients restricted to lie on

a fourth-degree polynomial with endpoints constrained

to equal zero.

For the Philippines, the following are the data

used in the study for the estimation of the St. Louis

Model: (in growth rates for the period 1986:Q1 to

2003:Q1, for a total of 69 observations)

1. Seasonally adjusted Gross Domestic

Product (GDP)

– previous estimations

under the St. Louis approach used the GNP

as the dependent variable. Since we only

measure the domestic implications and

effectiveness of fiscal and monetary policy,

an internal measure of growth is needed.

GDP is used over GNP so as to remove the

external factors (i.e., net factor income from

abroad). To correct the GDP series for

seasonality, the TRAMO-SEATS approach

was used.

2. Domestic Liquidity (M3) – it reflects a more

relevant feature of total money circulating in

the Philippine economy since it consists of

money supply, peso savings and time deposits

and deposit substitutes of deposit money

banks held by the general public. Quarterly

series of M3 is computed using the average

of each 3-month level. The M3 series were

sourced from the Bangko Sentral ng Pilipinas

(BSP).

3. Government Expenditures – it reflects the

high-level expenditures made by the national

government. Quarterly series of government

expenditures were obtained by adding up 3-

month levels of expenditures. The

Government Expenditure series were

sourced from the Bureau of Treasury (BTr),

since they have the sole responsibility of

releasing funds for the consumption of the

national government.

4. Exports – refers to all final goods going out

of the country, which is either classified as

domestic exports or re-exports. The Bureau

of Customs (BOC) should properly clear all

export goods. Exports series were sourced

from the National Statistics Office (NSO).

July 2004 Bangko Sentral Review

Following the Andersen-Jordan equation, the

modified version is formed as:

SAGDPGR =

αα

α +

ββ

M3GR +

ββ

GEGR +

ββ

XGR +

εε

where: SAGDPGR = seasonally adjusted GDP

M3GR = domestic liquidity

GEGR = government expenditures

XGR = exports

While the original version of the Andersen-Jor-

dan equation is formed as:

SAGDPGR =

αα

α +

ββ

M3GR +

ββ

GEGR +

εε

From the original and modified versions, the

polynomial distributed lag procedure was employed

with the lags varied from 8 quarters (2 years for the

medium to short term) and to 12 quarters (3 years

for the long term effect). Thus, it reflected a total of

four (4) estimation procedures. Statistical analysis

would be heavily dependent on the sum of lags, since

this has the interpretation of the long run effect of

domestic liquidity, government expenditure and

exports, to the seasonally adjusted GDP series.

Estimation Results

Modified St. Louis model estimation. Under

the 2-years (8-quarter lag) structure, the sum of lags

points to government expenditure having a long run

effect on GDP growth, assuming stationarity. Do-

mestic liquidity and exports do not have a long run

implication on GDP growth. The same result was

obtained under the 3-years (12-quarter lag) struc-

ture. Although both domestic liquidity and govern-

ment expenditure are statistically significant, the lat-

ter possesses a much higher t-statistic.

Original St. Louis model estimation. Under

the 2-years (8-quarter lag) structure, the sum of lags

reflects that government expenditure has a long run

impact on GDP growth, assuming stationarity. Under

the 3-years (12-quarter lag) structure, government

expenditure still possesses a longer lasting effect on

GDP growth by virtue of a higher t-statistic than do-

mestic liquidity (see Table 1).

2 YEARS LAG STRUCTURE

Coefficients

- M3 Growth Rate (M3GR)

- Gov’t Expenditure Growth Rate (GEGR)

- Exports Growth Rate (XGR)

T-Statistics

- M3 Growth Rate (M3GR)

- Gov’t Expenditure Growth Rate (GEGR)

- Exports Growth Rate (XGR)

3 YEARS LAG STRUCTURE

Coefficients

- M3 Growth Rate (M3GR)

- Gov’t Expenditure Growth Rate (GEGR)

- Exports Growth Rate (XGR)

T-Statistics

- M3 Growth Rate (M3GR)

- Gov’t Expenditure Growth Rate (GEGR)

- Exports Growth Rate (XGR)

* Significant at 95%

ORIGINAL

-0.00045

-0.11946

-0.00687

-1.75210

ORIGINAL

-0.12921

-0.31279

-2.75206*

-6.23533*

MODIFIED

0.02815

-0.11078

-0.00131

0.43897

-1.69588

-0.04508

MODIFIED

-0.14410

-0.29423

0.03927

-2.31017*

-6.71514*

1.87009

Table 1. Sum of Lags Summary Statistics: St. Louis Model Estimation

Bangko Sentral Review July 2004

Chow breakpoint test results. To bring about

the significance of any changes from 1994:3 to the

present - characterized by trade liberalization and the

Central Bank’s initial steps on implementing the

modified monetary targeting framework - testing for

structural stability was performed. Results show the

original model of the St. Louis equation proved to be

more stable than the modified version although both

models failed the stability testing as implied by a low

F-statistic, thereby higher p-values (see Table 2).

Overall, all the four equations reject the null hypothesis

of no structural change. Indeed, structural reforms

and adjustments made by the Philippine government

are deemed vital for sustainable growth in the

domestic economy.

Empirical Findings

Results from the econometric exercise show that

the policy conclusions of the original and modified St.

Louis equation are robust with respect to both the

specification of its lag structure and the imposition of

the polynomial restriction. From the econometric

exercise and the variations applied under the

Andersen-Jordan equation as well as the Almon

polynomial distributed lag procedure, we find a

counterfactual result, heavily dependent on the sum

of lags: that fiscal policy possesses long-run effects

on real growth rather than the standard monetarist

viewpoint. Thus, we accept the null hypothesis that

fiscal policy is more effective than monetary policy

in the long run. Given that the Philippines is only an

emerging market and that the basic foundations on

monetary and fiscal frameworks are still being fine-

tuned, it is likely that real growth is really influenced

by fiscal actions (budget management, tax collection,

revenue generation and investment projects) rather

than monetary actions (interest regulation, price

stability and financial soundness). The results were

further validated by the outcome of the Chow

Breakpoint and Granger Causality Tests – which

implies that the structural adjustments made by the

fiscal sector, and the expenditures made by the

national government are indeed vital for sustainable

domestic growth in the Philippines.

Reassessing the role of fiscal actions within the

framework of the St. Louis equation was further

emphasized. Referring still to the result of updated

estimates and tests, there is strong evidence validating

the idea that fiscal actions are significant in

determining GDP growth, once the effects of money

growth is taken into account.

Granger causality results. Under the

imposition of 4 lags in the Modified St. Louis model,

government expenditure growth Granger-causes

GDP growth and vice versa. But under the imposition

of 8 lags in the Original St. Louis model, only

government expenditure growth Granger-causes

GDP growth. Domestic liquidity and exports do not

Granger-cause GDP growth and vice versa for both

lag impositions and for both St. Louis models (see

Table 3).

EQUATION

ORIGINAL ST LOUIS

2 years lag structure

3 years lag structure

MODIFIED ST LOUIS

2 years lag structure

3 years lag structure

CALCULATED

F-STAT

7.75783

5.974042

3.155918

2.711809

REJECT

STABILITY

Table 2. Chow Breakpoint Test (1994:3 Breakpoint)

EXOGENEITY TEST

M (Granger causes) Y

1 year lag structure

2 years lag structure

Y (Granger causes) M

1 year lag structure

2 years lag structure

CALCULATED

F-STAT

1.26116

0.72986

0.63891

1.36085

REJECT

EXOGENEITY

OF POLICY

VARIABLE

YES

G (Granger causes) Y

1 year lag structure

2 years lag structure

Y (Granger causes) G

1 year lag structure

2 years lag structure

X (Granger causes) Y

1 year lag structure

Y (Granger causes) X

1 year lag structure

3.28916

3.38082

3.61739

1.57069

1.20956

0.53617

YES

Table 3. Granger Causality Test Results

July 2004 Bangko Sentral Review

Endnotes

This is a condensed version of the research paper submitted by the authors as partial fulfillment of the requirements for

Applied Econometrics at the De La Salle University last August 2003.

The use of M3 as intermediate target allows the BSP to prevent excess liquidity, which causes inflation.

Taken from data series of Dr. Josef Yap’s article on the output gap.

TRAMO, “Time Series Regression with ARIMA Noise, Missing Observations and Outliers”, and SEATS, “Signal

Extraction in ARIMA Time Series”, (Gomez and Maravall, 1996). TRAMO is a program for estimation and forecasting of

regression models with ARIMA errors and missing values. The program interpolates these values, identifies and

corrects for several types of outliers, and estimates special effects such as Trading Day and Easter and, in general,

intervention-variable type effects. SEATS is a program for estimation of unobserved components in time series follow-

ing the so-called ARIMA-model-based (AMB) method; the basic components are the trend-cycle, seasonal, and

irregular components, which are estimated and forecast with signal extraction techniques applied to ARIMA models.

The two programs are structured so as to be used together, both for in-depth analysis of few series or for routine

applications to a large number of them, and can be run in an entirely automatic manner. When used for seasonal

adjustment, TRAMO preadjust the series to be adjusted by SEATS. The two programs are intensively used at present

by data producing and economic agencies, including Eurostat and the European Central Bank.

Concluding Remarks

The shift towards an inflation targeting monetary

policy framework had enhanced BSP’s effectiveness

in its pursuit of price stability. The transparency

mechanism, the use of a wide set of information and

the forward-looking nature of inflation targeting all

contribute to a more effective monetary policy-

making. With the achievement of a stable price

environment, monetary policy plays a key role in

supporting a sustained growth path. However, results

from the St. Louis model show that the fiscal policy

component had driven economic growth from 1986:1

until 2003:1. In this paper, it is argued that shifts in

the level of aggregate demand cannot be readily offset

by monetary policy. Therefore, fiscal policy remains

a potent tool for offsetting major changes in the level

of aggregate demand, given floating exchange rates,

and some recent developments on inflation targeting

and budget management for the Philippines.

Halcon

The Authors

Mr. Neil Angelo C. Halcon is Research Analyst II of the Business

Expectations and Leading Indicators Sub-group, Economic and

Financial Monitoring Group of the Department of Economic

Research. He obtained his Master in Applied Economics (M-AE)

degree at the De La Salle University. He also obtained a Bachelor

of Arts degree in Economics with Computer Applications (cum

laude) at the San Beda College. Ms. Leah Melissa T. De Leon is

Senior Computer Operator II of the Corporate Planning Services

Group at the Government Service Insurance System (GSIS). She

obtained her Bachelor of Science in Commerce major in

Management of Financial Institutions and Bachelor of Arts in

International Studies major in European Studies (honorable

mention) at the De La Salle University. She recently completed

36 units of academic work towards obtaining a Master in Applied

Economics (M-AE) degree from the same university.

De Leon

Bangko Sentral Review July 2004

REFERENCES

Andersen, Leonall C. and Carlson, Keith M. 1970. A Monetarist Model for Economic Stabilization. Federal

Reserve Bank of St. Louis Review. April 1970.

Arestis and Sawyer. 2003. On the Effectiveness of Monetary Policy and Fiscal Policy. Economics Working

Paper No. 369. The Levy Economics Institute.

Baldacci, Cangiano, Mahfouz and Schimmelpfennig. 2001. The Effectiveness of Fiscal Policy in Stimulating

Economic Activity: An Empirical Investigation. November 2001. Preliminary Draft.

Barth, Marvin J. III, Remolona, E. and Wooldridge, P. 2002. Changes in Market Functioning and Central

Bank Policy. Bank of International Settlements.

Batten, D. and Hafer, R. 1983. The Relative Impact of Monetary and Fiscal Actions on Economic Activity:

A Cross-Country Comparison. Federal Reserve Bank of St. Louis Review. January 1983.

Batten, D. and Thornton, D. 1984. “How Robust are the Policy Conclusions of the St. Louis Equation?

Some Further Evidence”. Federal Reserve Bank of St. Louis Review. June/July 1984.

Boivin and Giannoni. 2003. “Has Monetary Policy become More Effective?”. NBER Working Paper No.

9459.

Gochoco-Bautista, M. 2001. “What Drives Monetary Policy?” Discussion Paper No. 0105. University of

the Philippines.

Gochoco, Maria Socorro 1993. “Are Money, Interest Rates, Outputs and the Exchange Rate Cointegrated?

Implications for Monetary Targeting”. The Philippine Review of Economics and Business. Vol.

XXX. No. 1, June 1993.

Guinigundo, Diwa C. 1999. Problems and Issues on Inflation Targeting: The Case of the Philippines. A

paper presented at the conference on “Framework for Monetary Policy in Post-Crisis Asia: Policy

Options and Inflation Targeting”, held on 8-10 December 1999 at Cebu City, Philippines.

Fu, D., Taylor, L. and Yucel, Mine K. 2003. Fiscal Policy and Growth. Working Paper No. 0301. Federal

Reserve Bank of Dallas.

Hafer, R. W. 1982. The Role of Fiscal Policy in the St. Louis Equation. Federal Reserve Bank of St. Louis

Review. January 1982.

Houben, Aerdt C.F.J. 1997. Exchange Rate & Monetary Strategy Options in the Philippines: The Search

for Stability & Sustainability. IMF Paper on Policy Analysis and Assessment. May 1997.

Jayaraman, T. K. 2002. Efficacy of Fiscal and Monetary Policies in the South Pacifc Island Countries:

Some Empirical Evidence. Fiji Islands.

Kimura and Kurozumi. 2002. Effectiveness of History-Dependent Monetary Policy. Bank of Japan.

King, Robert G. and Wolman, Alexander L. 1996. Inflation Targeting in a St. Louis Model of the 21

Century. Federal Rerserve Bank of St. Louis Review. May/June 1996. Vol. 78. No. 3.

July 2004 Bangko Sentral Review

Lane, Timothy D., et. al. 1999. IMF-Supported Programs in Indonesia, Korea and Thailand: A Prelimi-

nary Assessment. IMF Occasional Paper No. 178, IMF Washington, DC.

Levine, Rose and Renelt, David. 1992. A Sensitivity Analysis of Cross-Country Regressions. American

Economic Review. Vol. 82.

Meyer and Varvares. 1981. A Comparison of the St. Louis Model and Two Variations: Predictive Perfor-

mance and Policy Implications. Federal Reserve Bank of St. Louis Review. December 1981.

Paderanga, C. 2001. Recent Fiscal Developments in the Philippines. Discussion Paper No. 0110. University

of the Philippines. October 2001.

Poole, William. 2001. Monetary Policy in Uncertain Times. Remarks before the Missouri Bankers Associa-

tion. Senior Bank Management Conference. Mexico. January 2001.

Schaechter, A., Stone, M and Zelmer, M. 2000. Adopting Inflation Targeting: Practical Issues for Emerg-

ing Market Countries. IMF Occasional Paper 202. IMF Washington DC.

Tanzi, Vito and Zee, Hovell H. 1997. Fiscal Policy and Long-run Growth. IMF Staff Papers. Vol. 44.

Tetangco, Amando Jr. and Tuaño-Amador, Ma. Cyd N. 2002. Information Requirements of Inflation Target-

ing in the Philippines. Bangko Sentral Review July 2002. Vol. IV. No. 2.

Yap, Josef. 2003. Output Gap and its Role in Inflation Targeting for the Philippines. Bangko Sentral

Review July 2003. Vol. V. No. 2.