Chapter 5 Growth and Inflation in Malaysia

Manuel Guitián, and Robert Mundell
Published Date:
June 1996
  • ShareShare
Show Summary Details
Muthi Samudram

Malaysian economic growth since independence in 1957 can be divided into three phases. In the 1960s, when the economy grew at an annual average rate of 5 percent, Malaysia’s development base was associated with strong primary products, such as rubber and tin. These two commodities together accounted for 80 percent of export earnings, 35 percent of employment, and close to 40 percent of GDP during the period. Overdependence on primary products exposed the economy to external fluctuations. Some policy initiatives introduced then promoted import substitution where some small-scale industries began to be established. These developments were not sufficient to absorb the increasing labor force, and export promotion through the establishment of free trade zones began to be developed, as did a system of incentives. The 1970s witnessed the introduction of the New Economic Policy to improve the imbalances that existed in the distribution of wealth and employment among the people of the country. Export promotion was actively pursued, and, apart from the recession of 1985, Malaysia experiences continued growth.

Structural Adjustment in the 1980s

The 1980s were characterized by economic adjustments aimed at dealing with the prolonged commodity price slump, extensive government participation in the economy, and slow economic growth. The overall rate of expansion for the entire decade was only 4.9 percent. In the early 1980s, the Government undertook huge investments in heavy industries relating to automotive, cement, and iron and steel products. The Government also set up a large number of nonfinancial public enterprises with the aim of developing Malaysia into an industrial nation. After experiencing excessive foreign borrowing followed by a slump in economic activity, the Government embarked on a program of consolidation, liberalization, deregulation, privatization, and the promotion of foreign direct investment. The aim of these measures was to improve the financial position of the country and then to promote economic progress.

The recession of 1985 brought home forcefully the bad timing of the countercyclical strategy and the narrowness of Malaysia’s export sector in the 1970s and early 1980s. These strategies then caused the country’s external debt and debt burden to widen, and, with the implementation of the New Economic Policy, the Government’s overall deficit position began to worsen. These developments resulted in a six-year current account deficit in the balance of payments. Private investment was sluggish, and unemployment reached 8.5 percent in 1986. A structural adjustment program was implemented to correct the two emerging deficits. It included austerity measures that reduced public expenditures and promoted privatization. A policy package of incentives was introduced, and these measures succeeded in lowering the debt-service ratio and stimulating the economy.

With the assistance of large inflows of foreign direct investment, the economy began to surge ahead in the 1990s, averaging well above 8 percent growth in real GDP. Continued strong growth led by external demand posed problems for factors of production, with insufficient attention being paid to the supply side of the economy. Infrastructure bottlenecks and a lack of skilled labor have become problems for Malaysia’s continued growth prospects. This paper assesses the long- and short-term problems the economy is likely to face if the authorities pursue the objective of strong growth and the policy measures that are required to sustain reasonable growth with moderate inflation.

First, the potential output growth rate of the economy is estimated to provide a basis for analyzing the pace at which the economy can grow with price stability. Second, the wage-price mechanism is analyzed to identify the source of inflationary pressures and wage-price rigidities. Price rigidity helps determine how disinflation can be achieved and what sacrifices in output and employment must be made to achieve price stability.

Potential Output and Its Growth

From 1986 until 1994, the Malaysian economy grew significantly, well above 8 percent on average. As a result of the continued strong growth, unemployment fell from about 9 percent during the 1985–86 recession to 2.3 percent at the end of 1994. Between 1986 and 1994, employment in the economy increased at an average annual rate of 3.4 percent a year, while the labor force grew by 2.8 percent a year. Employment began to rise faster than the labor force beginning in 1987. As a result of the tightening of the labor market and the rise in capacity utilization, wages and prices began to accelerate. Inflation rose from 0.8 percent in 1987 to 3.7 percent in 1994.

Whether prices continue to rise depends on how fast the economy is growing relative to the growth of potential output. By definition, the potential output growth of an economy is the long-run growth path of output that is consistent with stable inflation. The level of potential output is determined by the long-term trend in capital formation and labor force growth together with the level of technological progress.

To estimate the level of potential output, we employ the production function approach, which directly links the economy’s potential output to the factors of production and technological progress.1 We estimated the following production function for the Malaysian economy for the period 1977–94:2

where Yt is the level of real GDP, Kt is the actual capital stock in 1978 prices, TV is actual employment, and 7 is the time trend. The cubic time trend (T3) allows the rate of technological progress to vary over time. The t -values are in parentheses.

We then estimated potential employment and trend capital stock and substituted them into the production function to obtain potential output. Potential employment is defined as trend labor force multiplied by (1-NAWRU). The trend labor force is obtained by running a regression of labor force against time trend.3 NAWRU is solved from the estimation of expectation-augmented Phillips curve. The sample time period is, again, 1977–94.


NAWRU is the equilibrium level of unemployment at which nominal wages grow at the same rate as labor productivity. Including NAWRU in the estimate of potential employment ensures consistency between labor market and product market equilibrium. Based on the above equation, NAWRU for 1977–94 was about 5.69 percent.

As shown in Table 1, the growth of potential output during the 1980s registered an average of 5.8 percent a year compared with the actual real GDP growth of an average of 5.7 percent. This suggests that the economy grew at capacity output level. In the 1990s, the average actual growth of real GDP has been 8.6 percent, which is higher than the average potential output growth of 8.4 percent (Figure 1).

Table 1.Growth of Potential Output and Contributors of Factor Inputs and Total Factor Productivity(Percent)
Contributions from Growth in


of Real GDP

Growth of

Real GDP




Sources: Ministry of Finance; and author’s estimates.
Sources: Ministry of Finance; and author’s estimates.

Figure 1.Potential Real GDP and Real GDP

(In millions of rinqqit)

If you exclude 1994, the average potential output growth for the 1990s is 7.9 percent. The increase in potential output in the 1980s is largely attributed to employment and capital stock with very little contribution from technological progress. In the 1990s, the contribution of total factor productivity to the growth of potential output has been significant.

The contribution of labor to potential output growth fluctuated between 1½ percent and 3 percent during the 1980s but has remained at about 2 –3 percent in the 1990s. The contribution of capital stock, however, fell from 4 percent in the early 1980s to1–1½ percent in the late 1980s. It began to rise in the 1990s owing to large inflows of foreign direct investment.

Short-Run Demand and Supply Nexus

In the short run, actual output deviates from potential output in response to shocks in nominal aggregate demand that lead to changes in prices and wages. In general, the aggregate short-run supply curve can be disaggregated into the reaction of prices to changes in unemployment as a result of changes in real output. The relationship of unemployment to output is the demand for labor, while the short-run Phillips curve describes the impact of unemployment on wages. The response of prices to wage changes is determined by markup pricing.

We analyze the split between inflation and output following Gordon (1990), where nominal GDP is divided into the two components of GDP deflator and real GDP.

where GDPN is nominal GDP, PGDP is the GDP deflator, and RGDP is real GDP. Taking the time derivative of (1), we have

which states that any change in nominal GDP(x˙) must be divided between a change in the aggregate price level (p˙) and a change in real GDP(y˙). Subtracting the long-run potential output from both sides of (2), we have


Equation (4) states that an excess of nominal GDP growth over the growth of potential output must be accompanied by some combination of inflation (p˙) and a deviation of real output from the potential.

In many episodes of recession and depression, a decline in output and employment appears to have constrained employees seeking more work and producers willing to produce more. Therefore, these episodes admit the possibility that actual output and long-run equilibrium output are two different and distinct concepts. If that is so, we can then relate price stickiness to excess nominal GDP and real output. Let us assume (Gordon, 1990) that a change of prices is always some fraction of excess nominal GDP demand movement; that is,

Then the remaining fraction of (1 - α) can be explained by the movements in real output.

The above equations suggest that an economy with relative price stickiness (a small α) must exhibit correspondingly large fluctuations in real output. But equation (5) ignores some crucial features of the economy, such as the effects of the Phillips curve, inertial effects, policy feedback, and supply shocks.

Price stickiness in equation (5) is represented by α, which denotes the marginal response of the rate of price change to a change in the excess growth rate of nominal GNP. But the Phillips curve relationship is defined in terms of price changes and the level of demand. Following Gordon (1990), the level of demand can be represented by unemployment or detrended output.

where Q^t is the log ratio of actual to potential output. A lagged inflation term can represent price inertia as follows:

Equation (7) is interesting not only in that price stickiness arising out of demand and lagged effects can be evaluated, but also because it shows how nominal demand can be disaggregated into price and output changes.

Equation (8) can be transformed into

The generalization of the Phillips curve hypothesis contained in (8) and (9) illustrates that price adjustment hypotheses can be expressed in several alternative forms and that the price change response to a change in nominal demand depends on the three parameters α, δ, and λ Equations (8) and (9) are estimated using annual observations from 1977 to 1994 and the results are in Table 2.

Table 2.Inflation and Price Stickiness in Malaysia

(Dependent variable = p˙t)

Sample period1977–941977–941977–941977–94
Note: p˙t = inflation,Q^t = log (actual real output as a ratio of potential output),x^t = excess of nominal GDP over long-run growth rate of real output, andq^t = deviation of real output from potential real output.
Note: p˙t = inflation,Q^t = log (actual real output as a ratio of potential output),x^t = excess of nominal GDP over long-run growth rate of real output, andq^t = deviation of real output from potential real output.

The estimates suggest that, for a given output gap, about 63 percent of an increase in the nominal demand relative to the potential output growth is absorbed by price increases within a single year. The relatively small coefficient of lagged inflation suggests weak price inertia. In fact, the insignificance of the coefficient indicates that prices were never sticky during the period. The estimates also suggest that about 15–16 percent of the output gap is reflected in the current inflation. Wage and price inflation cannot be sustained continuously if not accommodated by an increase in nominal demand. Figure 2 shows the growth of M2 and the growth of its velocity.

Figure 2.Growth of M2 and Velocity

(In Percent)

When nominal GDP growth began to accelerate in 1987, the velocity of circulation was the main contributor, but from 1989 to 1993, the growth in nominal GDP was accounted for entirely by monetary growth, with negative growth rates registered for velocity. For 1994, the developments of 1987 appear to have been repeated, with velocity once again contributing to the moderation of nominal GDP growth.

Wage-Price Mechanism

We need to evaluate aggregate wage-price behavior to determine the level of wage and price stickiness that exists in the economy. These relationships are relevant for establishing short-run price and output responses. We estimate the following:

where p˙te is the expected inflation. GA˙P is the growth of the gap between actual and potential real GDP. UL˙C is the growth in unit labor costs, while p˙mt is the percent change in import prices. The variables w˙t,un˙emt and pr˙odt are explained earlier.

The nominal wage equation represents union behavior in the bargaining process to maintain a desired real wage in line with the expected inflation rate. For a given inflation rate, the higher the labor productivity and the available opportunities for employment outside the firm, the higher the bargained wage. Outside opportunities can be captured by including a real wage [(w/p)t − 1] variable in the equation. A lagged nominal wage is included to represent wage rigidity arising from staggered collective agreements.4 In Malaysia, wage setting between unions and employers is not synchronized. A lagged wage variable reflects wage persistence that arises when workers determine their wages by looking at past wages.

The specification of the inflation equation allows prices to respond to excess demand in the goods market (proxy in the GA˙P) In addition, prices are expected to respond to unit labor costs. Import prices are an important component of the overall general price level, and changes in import prices are transmitted directly through the import of consumer goods or through the import of intermediate and investment goods via the cost of production (Figure 3).

Figure 3.Inflation and Gap (Actual GDP-Potential GDP)

Wage-price behavior depends on crucial parameters, first, the response of wages to prices or expected prices and the response of prices to unit labor costs. The parameters that reflect this behavior are a2 and b2. Wages and prices also respond to demand and supply conditions in the labor and product markets. Labor market conditions are reflected in unemployment (un˙em), while product market conditions are reflected in the GA˙P. The estimates of the wage and inflation equations are given below. Wages are represented by aggregate manufacturing wages, and inflation by the consumer price index:

The estimates of the wage equation indicate some inertia in the wage adjustment as suggested by the significant coefficient on the lagged wage variable. Basically, because wages are influenced by past behavior, present wages fully anticipate the actual inflation for the year. The estimated inflation coefficient suggests that a 1 percent increase in inflation leads to a 1 percent rise in nominal wages within a year. These responses appear to preserve real wages. The estimates also suggest that nominal wage changes respond to the current level of unemployment. In effect, this is a reflection of the new wage settlements rather than previous-year settlements because of the nature of the wage-setting prices. Labor productivity growth also influences rises in inflation. Given prices and the unemployment rate, a 1 percent rise in labor productivity leads to a 33 percent rise in wages. This sensitivity of wages to productivity provides another channel for wage flexibility that is independent of the level of unemployment.

The estimates of the inflation equation indicate that the coefficients of output gap, changes in unit labor costs, and import prices are all statistically significant. A 1 percent increase in the unit labor cost raises the inflation rate by 0.2 percent within a year. It appears that it is difficult to pass on cost increases to prices, particularly in an open economy like Malaysia, where competition is generated through import and export markets. Import prices are significant, and a 1 percent increase in import prices raises domestic inflation by a mere 0.2 percent. This suggests that the present growth structure, even without the effects of a gap, unit labor costs, and imported inflation, can generate a steady-state inflation rate of 3 percent in the economy.

The above analysis suggests that greater rigidity characterizes the response of inflation rather than of wages to shifts in demand conditions. Wages respond more fully to price changes but the response of prices to wages is limited. The response of inflation to the output gap is 0.49 percent, which suggests that changes in income growth have a moderate effect on inflation. The weak response of inflation to unit labor costs (0.16 percent) suggests that any change in the demand for labor as a result of an increase in aggregate demand will have a marginal effect on wages and prices.


To analyze the effects of disinflationary policy, the behavior of the wage-price mechanism is extremely important in determining how fast inflation can be reduced in the face of a loss of output. The Malaysian Government has recently committed itself to increasing government expenditures to finance infrastructure development, which is expected to continue until the year 2000. In the process, monetary policy is given the task of controlling inflation and inflationary expectations. Economic growth is expected to be generated not only by increases in development expenditures but also by continued inflows of foreign investment.

Disinflation can be achieved with only a small loss in output and employment if prices fall quickly in response to policy measures that are in place to reduce demand. In turn, nominal wages should adjust quickly to a drop in prices. Under the present circumstances, where there is excess liquidity in the economy, decreasing the money supply along with exchange rate appreciation is one disinflationary process that could be undertaken. Malaysia is an open economy with very little control over capital flows, and, therefore, interest rate hikes coupled with an appreciation of the currency can prevent inflows of private short-term capital that would otherwise generate growth in the money supply.

For the above policy combination to generate a lower sacrifice ratio, interest elasticity of the demand for money as well as the elasticity of domestic absorption with respect to real rates need to be low.


    AdamsCharles andDavidT. Coe1990“Systems Approach to Estimating the Natural Rate of Unemployment and Potential Output for the United States,”Staff Papers International Monetary FundVol. 37 (June) pp. 23293.

    GanW.B. andE.Robinson1994“Aggregate Supply and the Wage-Price Mechanism: Some Implications for Inflation Stabilisation in Malaysia,”paper presented at the Institute of Strategic and International Studies.

    GordonRobertJ.1990“What Is New: Keynesian Economics?”Journal of Economic LiteratureVol. 28 (September) pp. 111571.

    TaylorJohnB.1979“Staggered Wage Setting in a Macro Model,”American Economic Association Papers and ProceedingsVol. 69 (May).

    TorresRaymond andJohnP. Martin1990“Measuring Potential Output in the Seven Major OECD Countries,”OECD Economic Studies No. 14 (Spring).

This approach has been used by the IMF and the Secretariat of the Organization for Economic Cooperation and Development. See Torres and Martin (1990) and Adams and Coe (1990).

Similar estimates were carried out for the Malaysian economy. See Gan and Robinson (1994).

The trend labor force can also be obtained by using the Hodrick-Prescott filter to eliminate cyclical fluctuations.

Several articles discuss this issue; prominent among them is Taylor (1979).

    Other Resources Citing This Publication