Electronics and growth cycles in Singapore
TILA K ABEY SINGHE
Department ofEconomics, National University ofSingapore, 10 Kent Ridge Crescent,
Singapore 119260
E-mail: TilakAbey@nus.edu.sg
With the help of spectral and structural time series analyses it is found that the
demand for electronics in the USA goes through a cycle with a period of about
four to (R)ve years. Because of the synchronized nature of the electronics market,
the global demand for electronics is likely to follow a similar cycle. The electronics
cycle is unlikely to be caused by a regular business cycle. The major cause of the
electronics cycle appears to be a product cycle which results from the introduction of
new products to the market. As a result countries like Singapore which depend
heavily on electronics exports may face roller coaster rides, especially in their manu-
facturing growth. Product diversi(R)cation is the best solution to the problem.
I. INTR ODUCTION
After growing by 11% per annum between 1993 and 1995,
Singapore’s manufacturing sector entered an unexpected
downhill journey in 1996. The blame was borne by the
electronics sector. With about 45% of manufacturing out-
put and about 15% of GDP, the electronics sector was
accused for pulling down the growth of the Singapore
economy across many sectors. In fact, by mid-1996 the
pessimism gathered momentum propelling some to talk
about a recession.
1
Until Singapore was hit by the Asian Crisis in 1998, and
except for the 1985/86 recession, the Singapore economy
has not experienced the business cycles of mature econo-
mies. What Singapore has gone through is a growth cycle.
In a growth cycle, economic growth slows without substan-
tially aecting unemployment.
The other Asian NIEs have experienced similar growth
cycles. It appears that these growth cycles are closely
related to electronics cycles, caused by the uctuations in
global electronics demand. The objective of this exercise is
to (R)nd out whether there is a predictable electronics cycle
and how it relates to growth cycles of the Singapore
economy across various sectors. Moreover, if electronics
demand goes through cycles it is important to understand
what its cause is and what policy measures should be taken
to redress the problem.
II. METHODOLOGY
The major analytical tool used to examine cycles and their
inter-relation is spectral analysis. Many text books in time
series analysis provide a chapter on spectral analysis (see
for example, Fuller, 1996). In addition structural time ser-
ies analysis (Harvey, 1989) was also used for comparison
purpose. The Appendix provides a very brief account of the
methodology used in this exercise.
The presence of cycles are indicated by peaks in the
graph of the spectral density plotted against frequency.
The relationship between two series at dierent frequencies
can be analysed using cross spectral techniques. In the case
of structural time series analysis a model with a cyclical
component can be constructed in two ways, one with an
additive cycle and another with a cyclical trend. Both mod-
els provide an estimate of the frequency and the period of
the cycle if there is one. Although each of these methods
has its own strengths and weaknesses we use both methods
to see whether they are in agreement regarding the cycles.
Applied Economics ISSN 0003±6846 print/ISSN 1466±4283 online # 2000 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
Applied Economics, 2000, 32, 1657±1663
1657
1
This exercise was carried out in 1996 when there was wide interest on the topic as the electronics sector appeared to be taking a nose
dive.