Employment Equation:
13 Although it is possible that the real minimum wage varies over time, the fraction affected value obtained from the difference between the old and new real minimum wage is not appropriate because
the variation in the fraction affected is not only driven by the variation in the minimum wage but also by the variation in the price index (the denominator used to transform the minimum wage in the real terms).
14In addition, in order to capture any measurement error, similar to Lemos (2004d), the fraction at and
for less than one year15. As pointed out by Manning (2003a), the minimum wage in
Indonesia seems not to exactly act as the wage floor, but it is more likely to be an
instrument for raising the workers standard o f living. Not surprisingly, the fraction at
in Indonesia is only around 4% o f total paid employment, while in Brazil it is 12% o f
total employment.
On the other hand, the “fraction below” measures the degree o f non-compliance with
the minimum wage. Although the minimum wage policy legally covers all o f the paid
employment in Indonesia without exception, some workers remain paid below the
minimum level, particularly in rural areas and small enterprises where labour unions
are not effective. This condition is also combined with the lack of enforcement
capabilities by the government and ineffective sanctions for those employers who do
not comply with the policy. A higher minimum wage level might indicate an increase
in the proportion o f workers paid below the minimum wage level because workers
who experience a job loss after the minimum wage increases are likely to accept
wages below the minimum wage level rather than become unemployed in the absence
o f unemployment benefits. On average, the fraction below in Indonesia is 22% o f total
paid employment.
15 I f th e d u ra tio n o f w o rk is m o re th an o n e y e a r, e m p lo y e rs s h o u ld p a y th e sa la ry o f th e ir m in im u m w 'age w o rk e rs m o re th a n th e m in im u m w a g e lev el, a lth o u g h it is n o t c le a r b y h o w m u ch .
5.4. Data Sources
The main source o f the aggregate data used in this study is the Indonesian Labour
Force Survey (Sakemas). It is a regular labour force survey in Indonesia conducted by
the National Central Bureau o f Statistics (BPS) annually and/or quarterly since 1986,
except in 1995 when BPS conducted the Intercensal Demographic Survey (SUPAS)16.
In practice, SUPAS 1995 covered the same questions as Sakemas in the labour force
section with a sample which was three times bigger than the common Sakemas. The
main objective o f Sakernas is to estimate and monitor the labour force statistics and
characteristics in Indonesia. This survey provides a rich source o f the cross-sectional
labour force data, covering about 160,000 respondents (about 0.1% o f population)
each year. Relating to the objective o f this study, the individual data used in this study
is aggregated at provincial level.
As pointed out by Smith et al (2002), compared to the other countries’ labour force
surveys, there are two main limitations o f the Sakemas. Firstly, this survey is a cross-
sectional labour force survey. Therefore, there is no information about the longitudinal
change in the labour force across individuals. As a result, respondents might be
different across years. Secondly, relating to the earnings data, this survey does not
cover the self-employed (informal sector) earnings before 2002. The only earnings
data available before 2002 are for respondents with paid employment status.
However, in accordance with this chapter’s objective, this survey is powerful enough
to capture most o f the changes in the labour force characteristics across provinces. As
noted above, this chapter uses survey data from 1989 to 2003 across 26 provinces in
Indonesia. The reason for using data from 1989 is the fact that the minimum wage
began as an important policy in the Indonesian labour market in the late 1980s.
Following a similar format to Suryahadi et al (2003), the panel data set is only
constructed for 26 provinces in Indonesia. To simplify, this panel data set does not
cover some new provinces that have been formed in 2001 during the decentralization
era. In practice, there have been 33 provinces in Indonesia since 2001, including seven
newly formed provinces being added to the former ones. In this case, data from the
newly formed provinces are combined with those o f their original provinces. In
addition, East Timor data is excluded because o f its independence from Indonesia in
1999.
Based on Sakemas, the employment status is divided into five different categories,
including (a) self-employed; (b) self-employed and assisted by non-permanent
employees; (c) employer and assisted by permanent employees; (d) workers or paid
employment; (e) family or unpaid workers. Although the self-employed is the largest
group o f employment which contributes to the informal sector, this study specifically
focuses on the workers or paid employment category as the sector o f employment
legally covered by the minimum wage. The effects o f the minimum wage on the other
employment categories, including employment in the uncovered sector, will be
specifically discussed in the next chapter.
The other crucial data used in this chapter are obtained from different sources. The
and the BPS publications17. The provincial consumer price index data which are used
to transform the minimum wage and the average wage variable in real terms are
obtained from the BPS publication.