Eko Atmadji Fakultas Ekonomi.
Universitas Islam Indonesia e-mail: ekoatmadji@gmail.
Abstract The administered price of fuel has increased quite frequently since 1970.
It has raised a concern on the survival ability of manufacturing firms that could be weakened by such policy.
This research objective is for confirming whether the policy of increasing fuel price affects the survival of manufacturing firms in Indonesia.
By using an Instrumental Variable Probit Model, it demonstrates that increasing fuel price does not affect firmsAo survival in the short-run, middle-run, as well as longrun.
Nevertheless, the z-statistics increase when the period of estimation is expanded.
It means that if the time span of sample observation is widened, the result might be different.
Keywords: Fuel price, survival ability, instrumental variable probit model JEL classification numbers: H27.
D22 Abstrak Harga bahan bakar minyak (BBM) selalu meningkat sejak 1970.
Kenaikan harga ini telah menimbulkan kekhawatiran pada kemampuan kelangsungan hidup perusahaan manufaktur di Indonesia yang bisa dilemahkan oleh kebijakan tersebut.
Penelitian ini dilakukan dengan tujuan mengonfirmasi apakah kebijakan kenaikan harga BBM mempengaruhi kelangsungan hidup perusahaan industri manufaktur di Indonesia.
Dengan menggunakan Instrumental Variable Probit Model, hasil penelitian menunjukkan bahwa peningkatan harga BBM tidak mempengaruhi kelangsungan hidupperusahaan dalam jangka pendek, menengah, serta jangka panjang.
Namun demikian, z-statistik meningkat ketika periode estimasi diperluas.
Ini berarti bahwa jika rentang waktu pengamatan sampel melebar, hasilnya mungkin akan berbeda.
Kata kunci: Harga BBM, kemampuan bertahan hidup, instrumental variable probit model JEL classification numbers: H27.
D22
INTRODUCTION
In Indonesia, the price of fuel is determined by government with a low price policy.
Usually, the price is lower than the average cost of fuel production, and even lower than the fuel prices in the countries of Southeast Asia with the exception of Brunei Darussalam.
Nevertheless, it is difficult to maintain low fuel price due to increasing of fuel subsidy burden.
The rise of world oil price and increasing fuel domestic consumption cause the escalating of fuel For that reason, government has constantly raised fuel price in the last 40 Every announcement of raising fuel price always heated political and social situation because such policy intiates inflation.
Manufacturing industry sector has also been taken the effect of raised fuel The multipiler effect of increasing fuel price influences the rise of input prices and accordingly causing higher production 1 There are two options in adjusting Fuel consumption for the energy sector of the manufacturing industry is average on 30.
85% (Indonesian Energy Sta- Fuel Price Increase and ManufacturingA (Atmadj.
the effect of cost burden rise on production operation, making production more efficient or shuting down the business.
make production more efficient will sacrifice some inputs including the workers.
promotes higher unemployment rate.
shutting down the business is the choice for many firms, the massive closure will reduce national production and also rising After 2000, fuel prices had been increased for five times.
During 2002-2006, there were three times of fuel price increasing2.
The most phenomenal of raising fuel price was in 2005 which was increased by more than 100%.
Several fuel price increases repeatedly was presumed brought bad impact to firms in the manufacturing It weakened the survival of manufacturing firms.
It is assumed that manufacturing firms are misery in facing the rise of fuel Based on the assumption, this study question is how big the effect of rising fuel prices on the possibility of manufacturing firms exit from the market.
This study will use survey data medium and large firms in manufacturing industries, published by BPS with the period 2002-2006.
This period was elected because in this period fuel price increased three times.
It is expected to observe the impact of fuel price increases on the survival of the firm directly.
Table 1 shows that from 2002 to 2006, the firm mortality rate are 5.
1%, 7.
6%, 6.
9%, and 12.
2% of its previous year, respectively.
The trend showing arising percentage of mortality rate indicates the problems become more difficult tistics 2.
The cost of fuel consumption of a firm contributes less than 5% of total cost of production on the average with a minimum share of close to 0% and the maximum share of more than 50% (Ministry of Energy and Mineral Resource.
Republic of Indonesia, 2.
The fuel price increase occurred in 2002 .
rom Rp1,450,- to Rp1,550,- for a premium and of Rp1,200,- to Rp1,110,- for a diese.
, 2003 .
rom Rp 1,550,- to Rp1,810,-for premium and from Rp1,110,- to Rp1,650,- for a diese.
, and 2005 .
rom Rp 1,810, - to Rp4,500, - for a premium and of Rp 1,650, - to Rp4,300, -for diese.
to be faced.
While, the number of new firms that entered from 2002 to 2006 respectively also increased, which are 919, 1041, 2036, 1562, and 11,579 firms.
demonstrates that manufacturing industry in Indonesia is quite attractive.
In the period of transition from a multidimensional crisis conditions experienced by Indonesia since 1998 leading to a more stable economic conditions.
Although the number of firms were leaving the market rises but new firms were coming up as well.
Table 2 shows that the firms which were 5 to 10 years old have the lowest mortality rate .
1%).
Meanwhile, the highestmortality was a firm with 1 to 10 years old .
9%).
For a firm with over 10 years old also had a high mortality rate .
73%).
Based on these figures, the firm whose age over than 10 years have problem continuing its business.
While new entrances have not had much experience in managing From the ownership point of view, firms owned by national private had the lowest mortality rate in the amount of 19%.
Meanwhile, state-owned and foreignowned firms had a mortality rate of 34.
84% respectively.
It can be said that, national private owned firms had larger power struggle in maintaining their business than foreign-owned or stateowned firms.
Medium-large firms had the lowest mortality rate .
57%).
While large and medium firms had a mortality rate of 64% and 20.
49% respectively.
Large and medium manufacturing firms had higher mortality rate than medium-large firms.
Later, firms with export orientation had a mortality rate that was smaller than that not export-oriented firms were which 1% versus 20.
Therefore, the firm that last longer were firms with export orientation.
In terms of the use of fuel in the production process, the firm which had more than 50% fuel costs share to total production costs had high mortality rate.
However, the highest mortality rate was the firms which have fuel costs share top roduction costs between 30% to 40% in the amount of 24.
68% (Table .
The lowest mortality rate in manufacturing firms was the firms which have fuel cost share between 10% to 20% which is 16.
This figure is lower than the mortality rate of the firms which have less than 10% fuel cost From this phenomenon, it can be assumed that the effect of fuel price increases are inconsistent with the firm's demise.
Table 1: Numbers of Firms in Manufacturing Industry Sector
Year
Numbers of Survived 21,392 21,146 20,324 20,685 20,729 29,468 Firms Numbers of Closed Firms Oe
1,165
1,863
1,675
1,518
2,840
Numbers of New Firms 1,041 2,036 1,562 11,579 Oe Dead Percentage (%) Source: The Annual Survey of Medium and Large Enterprises.
Central Statistical Agency Table 2: The Anatomy of Dead and Survived Firms Manufacturing Industry Sector 2002-2006
Dead (%) Survived (%) Numbers of Firms Based on Ownership
National Private 25,654 Foreign
2,078
Government Owner Based on Size Middle .
- 100 worker.
21,566 Big Middle (>100 Ae 500 worker.
5,226 Big (> 500 worker.
1,867 Based on Age
1-5 year
7,249
> 5 -10 year
5,863
> 10 year
15,547
Export Oriented No Export
24,274
Export
4,385
Source: The Annual Survey of Medium and Large Enterprises.
Central Statistical Agency Table 3: Firms Description Based on Fuel Cost Share Group Dead Survive Numbers of (%) (%) Firms 1: < 10% 25,826 2: 10% x < 20% 1,784 3: 20% x < 30% 4: 30% x < 40% 5: 40% x < 50% 6: 50% Source: The Annual Survey of Medium and Large Enterprises.
Central Statistical Agency Fuel Cost Share Group (X) Fuel Price Increase and ManufacturingA (Atmadj.
METHODS The possibility of failure of firm to survive ini a market can be denoted as:
p (T O t ) = F .
) .
where T is survival spell and t is observation spell.
T t means a firm leaves a market before and the same of the end of observation spell.
While F.
is failure function.
It means that:
p (T O t ) = 1 Oe F .
) = S .
) .
Equation .
is called survival equation which is the possibility of a firm to survive in a market during the observation.
Simply, it is likely the firm can not survive because the market price is lower than the point of shuting-down.
The failure of a firm to lower its shut down point is caused by many internal and external factors.
According to Jovanovic .
, the characteristics of a firm in the face of market dynamics will determine the firm's ability to survive in the competition.
The main factors of survival are age and firm size.
The older the firm the more experienced the firm to deal with many obstacles.
The larger the firm the greater the chances of survival because it has achieved economies of scale.
According to Ghemawat and Nalebuff .
, the propositions of Jovanovic model about the size of a firm can be wrong if the firm faces a receding market.
The bigger the firm will be the greatercosts than other firms which are smaller size.
Demand absorption of bigger firmAos product become smaller and, therefore, it is more difficult to survive for bigger firms.
Based on two contrasting models, many studies refer to both models.
By focusing on the age and size of the firm, the researchers found that the age and size of the firm is directly proportional to the abil- ity to survive3.
However, other results4 found that the size and age did not affect the survival of the firm.
The result of studies are consistent with GibratAos law5.
Further development is not only focused on the size and age of the firm itself but adds other factors both other firm characteristics or external factors.
Moel and Tufano .
and Perez et al .
included financial elements as determinants of firm Factors technological advances affect the survival of the firm is the conclusion of research.
Cefis and Marsili .
, and Buddelmeyeret et al.
A more comprehensive study conducted by Bernar and Jensen .
Findings showed not only a factor of the size and age of firms that affect the survival of the firm but it was also productivity, export orientation, and specialization.
Another factor that reduces the survival of the firm is the ratio of non-production workers, the number of factories more than one, and the activities of multinational corporations.
FirmAos survival researches in Indonesia were conducted by Narjoko and Hill .
, and Kuncoro .
Narjoko-Hill analyzed the surivival of the firms in the middle of multidimensional crisis of 1998.
The findings were foreign-dominated and export-oriented firms had a greater probability of survival than others.
Kuncoro research focused on the effect of regional autonomy to the firmAos survival.
The regional autonomy was proxied by the intensity of collecting bribes by local authorities to the firm.
Kuncoro found that the intensity of official visits by local government officers to the firm increased the probability of firms to exit the market.
Among them are Goddard et.
Koening and Faggio .
Harris and Trainor .
, and Johansson .
Singh and Whittington .
Hardwick and Adam .
Chen and Lu .
, and Lensink et.
GibratAos law said the probability of a given proportional change in size is the same for all firms in a given industry, regardless of their size at the beginning of the time period (Mansfield, 1.
This research will incorporate fuel price factor as the main factor that affect survival of manufacturing firms.
Control factors over fuel price variable are a firm characteristic factors and demand factors.
Demand factors are incorporated into the control factor because of rising fuel prices affect the survival of the firm through the supply side.
If it is not offset by the demand side, uncontrolled behavior of error term will be more difficult to handle.
The econometric model of this research is as follows.
p (T > t ) = bX cY dZ e p(T > .
is dependent variable which has only two values, one if a firm still survive until the end of the observation period and zero if other.
X is a firm characteristics vector.
Y is a demand vector, and Z is main factor which is fuel price.
Meanwhile, b, c, and d are coefficient vector of estimation.
Firm characteristics and demand factors include the following.
Firm Characteristics Factors The variables representing factors firm characteristics are firm size, firm age, firmproductivity, and ownership status.
Firm size uses amount of labors as a The labor is directly related to production.
Firm age is determined by age in 2002, the firm at the commencement of the study.
Productivity are measured by the ratio of the volume of output and employment in the production.
The volume of output is obtained from dividing the value of output with price index adjacent to the commodity ISIC group produced.
Ownership status is divided into three groups which are foreign ownership, domestic private ownership, and government ownership.
The firm is considered foreign owned if there is a foreign element in their possession though not dominant.
It takes two dummy variables in explaining ownership status.
The value of foreign owner- ship dummy variable is one if there is a foreign element and zero if other.
The value of domestic private ownership dummy variable is one if the firm is worth a hundred percent privately owned nationwide and zero if other.
Demand Factor Variables representing the demand factor is the orientation of exports and demand Export orientation was chosen as one of the variables that represent the demand for exports is part of the demand for industrial manufacturing goods.
While demand growth variable is used as a variable that represents the request because it directly has a variable demand itself.
Variables such as export orientation dummy variable with a value of one if there is a partial or total sales are exported and the value of zero if other.
While demand growth is volume output use variable.
Main Factor The main variable of this research is fuel The use of fuel prices as a independent variable will lead to another problem, the firmsAo reaction of fuel price increases It depends on the size of the share of the fuel cost in cost of production.
If the price of fuel is the main variable, it will lead to biased behavior.
Therefore, the fuel price variable is replaced by a variable fuel cost share toward the cost of production.
Replacement of this variable does not affect the analysis because Atmadji .
has shown that the share of fuel cost variable is a monotonic transformation of the variable fuel prices.
Using this new variable has two First, it represents the movement in fuel prices.
Second, it captures the fuel usage variations at each firm.
The econometric function to be estimated is:
p (T > t ) = f.
hare, age, size, productivity, foreign, private, export, growt.
Fuel Price Increase and ManufacturingA (Atmadj.
Share is the fuel costs share to the cost of production, age is the age of the firm, size is a measure of firm size, productivity is the productivity of the firm, foreign is foreign ownership, private is private ownership, export is export-oriented, and growth is the growth in demand.
Estimation will use the Panel Probit method.
Probit estimation method requires only two points in time.
Hence, the estimation uses six sets of time periods to accommodate the three-time fuel prices increases in the period 2002-2006.
Three sets of time which is 2002-2003, 2003-2004, and 2005-2006 represents a short-term effect.
For medium term, the sets of time are 2003-2005 and 2004-2006.
A time set for long term effect is 2002-2006.
RESULTS
The estimation has two problems.
The first problem is the presence of heterogeneity in the data to be estimated.
Heterogeneity is caused by large variations of the firms that are in nine sub-sectors of the manufacturing industry.
Treatment of heterogeneity is done by giving a dummy variable for each sub-sector of manufacturing.
Dummy variable D-food .
rom the food and beverage sub-secto.
will be equal to 1 for food and beverage firms, and zero if other.
Other dummy variables.
D-textile .
extile sub sector.
D-wood .
he wood processing sub secto.
D-paper .
aper industry sub secto.
D-chemical .
hemical goods sub secto.
D-cement .
ement sub secto.
Dmetal .
etal industry sub sector.
Dmachine .
rom sub sector machine industr.
also include the value of 1 on the firms that are in the sub-sector in accordance with the group.
Outside the group were given a value of 0.
The anchor sub sector is miscellaneous sub sectors.
The second problem is that there are elements of endogeneity in main independent variable, fuel share variable.
Rising fuel prices will increase the share of fuel costs in production cost.
On the other hand, ris- ing fuel prices also cause the price increase of other inputs hence it increases the cost of other inputs.
The cost increasing of other inputs will reduce the cost share of fuel Therefore, instrumental ariablefrom probit procedure applies.
There are three variables selected as an instrument which are the ratio of total corporate tax contributions on a district by district nominal GRDP (Gross Regional Domestic Produc.
, the average age of the firm at the same districts, and the average size of firms in the same district.
The result of application of instrumental variables method in panel probit estimation is presented in Table 4.
The survival probability calculations based on the estimation presented in Table 4 with the average value of each of the independent variables .
xcept the dummy variable.
is in Table 6.
Survival probability calculation results are grouped according to the share of the fuel cost at each firm.
Marginal effect of the estimation result of Table 4 are presented in Table 5.
Calculating the marginal effect using the method described by Anderson and Newell .
Sub-sectoral Characteristic Influence The estimations (Table .
show that dummy variable D-textiles and D-wood are significant and positive in the long term.
means that firms from textile and wood sub-sector had greater possibility to survive compare to other manufacturing firms from other sub sectors.
While in the mediumterm period .
3-2005 and 2004-2.
showed that all sub-sectors of the manufacturing industry had no difference in survival probability because all sub-sector dummy variables are statistically insignificant.
In the short term, only the period 2002-2003 showed the differences in survival p robability for the food sub-sectors and the textile sub-sector compared toothers.
In other short-termperiod, all subsectors had the same survival probability.
In general, the survival probability was no different to all sub-sectors except the survival probability in the long term.
In this long term period, firms in textile sub sector and wood sub sector showed better performance compare to firms in other sub Presumably, there was an element of high performance in export which leads to the difference between the two subsectors and other sub-sectors in terms of survival probability.
Sub-sectors of textiles and wood had high export performance since most of the exported products had very high demands for Indonesian textiles and furniture products from abroad.
Table 4: Estimation Result Using Probit Instrumental Variable Variable FuelShare ln_size ln_Age D-private D-foreign D-export D1-food D2-textile D3-woods D4-papers D5chemical D6cement D7-metal D8machine Constant Wald Chi2 Numbers of Observation (-1.
33E-05
***
75E-09
(-0.
***
***
(-1.
(-2.
(-2.
(-1.
(-1.
(-0.
(-1.
(-1.
***
26***
34,149
Time Period
2003-2004 2005-2006 2003-2005 2004-2006
(-1.
(-1.
8E-05
3E-05
(-3.
***
***
(-2.
***
***
***
***
***
***
73E-08
45E-08
79E-08
53E-08
(-1.
***
***
***
***
***
***
(-2.
***
***
(-3.
***
(-1.
(-0.
(-0.
(-0.
(-1.
(-0.
(-0.
(-0.
(-.
(-1.
(-1.
(-0.
(-0.
(-0.
(-0.
(-0.
(-0.
(-0.
(-1.
(-0.
(-1.
(-0.
***
18*** 1216.
04***
48*** 1967.
34***
33,034
34,140
34,060
32,524
Notes: ***: significant at significance level 1%.
**: significant at significance level5% *: significant at significance level 10%.
Numbers in the parentheses is Z test (-1.
***
39E-08
***
***
***
***
(-0.
(-2.
***
(-3.
***
(-0.
(-0.
(-1.
(-1.
86***
33,279
Fuel Price Increase and ManufacturingA (Atmadj.
Tabel 5: Estimation Result Using Probit Instrumental Variable Variable FuelShare ln_size ln_Age D-private D-foreign D-export D1-food D2-textile D3-woods D4-papers D5chemical D6-cement D7-metal D8machine Wald Chi2 Numbers of Observations Marginal Effect (-0.
E-06
***
E-10
(-0.
***
***
***
(-1.
(-1.
(-2.
(-1.
(-1.
(-0.
(-1.
(-1.
98***
34,149
Marginal Effect (-0.
E-06
(-1.
***
E-09
(-1.
***
***
(-1.
***
(-1.
(-0.
(-0.
(-0.
(-1.
3***
33,034
Time Period
2005-2006 2003-2005
Marginal Marginal Effect Effect .
E-06
*** .
***
*** .
***
E-09
E-09
***
*** .
***
***
(-2.
*** .
***
(-0.
(-0.
(-0.
(-0.
(-1.
(-0.
(-0.
(-0.
(-0.
(-0.
08***
83***
34,140
34,060
Marginal Effect (-0.
***
***
E-08
***
***
***
***
(-0.
(-0.
(-0.
(-1.
(-0.
(-0.
(-0.
19***
32,524
Marginal Effect (-1.
***
E-08
***
***
***
***
(-0.
(-2.
(-2.
***
(-0.
(-0.
(-0.
(-1.
(-1.
92***
33,279
Notes: ***: significant at significance level 1%, **: significant at significance level5% *: significant at significance level 10%.
Numbers in the parentheses is Z test Demand Influence The presumption that exports triggered the improvemant of survival probability were supported by the results of the dummy variable D-export which were significant in almost all periods of short, medium and long term, unless the medium period 20042006.
The difference between the survival probability of firms which was not exportoriented and export-oriented is between 1 percent and 3.
2 percent (Table .
The difference was not much but was able to make the firms more competitive and better survival probability than firms which are not export-oriented.
Demand growth variable was significant in the short and medium term.
For the short term, this variable was significant in 2003-2004 and 2005-2006.
While in the medium term, it was significant for the periods 2003-2005 and 2004-2006.
The same result occured in the marginal effect of this variable (Table .
In the period 2002-2003 and the period 2002-2006 .
ong-ter.
, demand growth was not significant.
In the transition period of crisis, variable demand growth was not as important as other variables to affect the survival probability.
While in normal times, it started an important variable for the existence of the firm.
In the long run, the influence of the transition process was too strong for the firm's survival probability that this variable is not significant in general.
FirmsAo Characteristics Influence For dummy variables such as D-foreign .
o element of foreign ownershi.
and D-private .
% domestic private ownershi.
showed a different level of significance.
For Dforeign variable was significant period 20032004 .
hort ter.
, 2004-2006 .
edium ter.
, and 2002-2006 .
ong-ter.
While that was not significant to the variable D-foreign in the period 2002-2003, 2005-2006 .
econd ter.
, and 2003-2005 .
edium ter.
This means that there was an element of foreign ownership did not affect the increase in survival probability for the short and medium While the D-foreign variables that were significant in the short term was negative, which means the existence of foreign elements actually increase the possibility of the firm to exit from market.
As for the periods 2004-2006 and 2002-2006, the D-foreign variable showed a high level of significance.
If it is viewed from the marginal effect level (Table .
, the variable D-foreign only had an impact in the medium term .
and long term .
, which means the difference between survival probability ownership firms with no foreign element that was foreign elements was 4.
55% .
and 65% .
ong-ter.
Different result occured in national private ownership.
This variable was convincingly significant for all periods .
ee Table .
However, in terms of marginal effect level (Table .
, the difference between a national privately owned that was not happening in five periods with an insignificant period .
The differences in survival probability were for the five periods between a 100% privately nationaly owned firms and others between 1.
68% to 5.
(Table .
Thus, national private ownership firms could improve the survival ability of the firms from bankruptcy.
This is because national private firms owners knew betterhow to deal with domestic issues, economically and politically.
In the long run, productivity variable showed a significant performance in the impact on the survival probability of the firms from 2002-2006.
In the medium term, productivity was only significant in the period 2004-2006.
While in the period 20032005, productivity was not significant.
the short term, productivity was not significant in the periods 2002-2003 and 20032004.
While in the period 2005-2006, productivity was important factor for survival in the market.
The pattern shows that productivity was not that important in the transition period from the multidimensional crisis into the normal condition, as firms strived to survive by not raise the productivity.
However, after the transition was complete, productivity began to be important factor in its influence on the survival of the firm.
In general, productivity was still important to the survival of the firm because in the long run the performance of productivity is significant.
For other firm characteristics, age and size were dominant for all periods.
Both Table 4 and Table 5 shows that the two independent variables were highly significant in influencing the survival probability ofmanufacturing firms.
Age had a major impact on the survival of the firm.
The older the firm, the higher the ability of the firm to Fuel Price Increase and ManufacturingA (Atmadj.
cope the problems in competition and internal or external distractions.
From Table 5, it is showed the marginal effect of the variable age on survival probability of the firms.
These figures show that every 1% increase in the age of the firm would increase the chances of firms survival from 1.
72% until In the short term, this figures of marginal effect increase from 0.
, 0.
, until 0.
While in the medium term, the period 2003-2005 had lower marginal effect rate than in the period 2004-2006.
While in the long term .
, the marginal e ffect rate was close to the period 2004-2006.
After the transition was complete and enter the normal time, the age factor became more important for firms to improve the survival probability manufacturing firms.
Although the firm age had the same performance with firm size, it had smallermarginal effect.
In the short term, marginal effect of firm age in the 2002-2003 was greater than the marginal effect of firm age in the period 2003-2004.
However, in the 2005-2006, firm age had the greatest marginal effect which is 0.
It means that every increment of one percent the size of the firm would increase the survival probability of 2.
Based on the value of marginal effect, an increase of firm survival probability during the period 2004-2006 .
edium ter.
had higher probability of survival than those faced by firms in the period 2003-2005.
It is likely that at normal times, the size of the firm was becoming increasingly important in the challenge of external and internal problems.
This suspicion was even stronger when the estimation results indicated it in the long term.
Firm size variable had higher marginal effect value than the marginal effect of the short and medium term.
It means that firm sizewas an important variable in dealing with the problems of the firm.
Large size firms were difficult to be shaken by numerous problems since most of their production scale had reached the economies of scale.
In addition, the significance of firm size variable was additional evidence how Gibrat law does not apply to the manufacturing industry in Indonesia.
Main Factor: Fuel Cost Share Figures from the estimation of fuel cost share were negative except for 2005-2006 and This is consistent with the notion that the increase of fuel cost share would raise the possibility of firmAos death.
What happens in 2005-2006 and 2003-2005 showed that the increase of fuel prices at those time did not affect the survival of the firm.
Based on Table 4 and Table 5, the fuel cost share variable was statistically insignificant to influence the survival probability of a firm, both in the short term, medium term, or long term.
However, in z test point of view, it shows that the longer the time spell observed would increase the value of its z test.
It subtly suggests that the increase in fuel prices remained miserable In the short term, the firm was still able to adjust due to rising fuel prices.
But if it continued to face, rising share offuel costs would weaken the survival of the firm.
From the survival probability point of view, there was low possibility to die for manufacturing firms because the possibility was above 95% on average (Table .
6 In the short term, the total survival probability of firms was above 96%.
In the medium term, the lowest survival probabilities were 11% .
25% .
While in the long term .
, survival probability of manufacturing firms 27% on average.
If related to the fuel cost share in the survival probability, there were inconsistencies in the results.
The firm with the highest share ofthe cost of fuel did not necessarily have the lowest probability of survival.
Instead, the firm with the lowest fuel cost share, did not necessarily have the Table 5 exhibits survival probability calculation based on Table 4 and the mean of each independent variables.
highest probability of survival.
Table 6 shows the sequence of firm survival probability in each period.
The table shows thatthe inconsistency occurred in the short term, medium term and long term.
In the short term, the survival probability of the order of lowest to highest were inconsistent when compared to the three short term periods.
The firm with the lowest fuel costs share on the average was not the firm with the highest survival probability.
While firms with the highest share of the cost of fuel did have the lowest survival rate probability unless the period 2005-2006.
In the medium term, the order of survival probability from the lowest until the highest were not necessarily the same as the order of fuel cost share from the highest to the lowest level.
Survival probability was a random sequence.
Even for the period 2004-2006, the firm with the lowest share of fuel cost .
ess than 10%) appeared to have the second lowest survival probability after the firms which have the share of fuel costs between 40%-50%.
Firms with the highest share of fuel costs .
ore than 50%) were in the third lowest order of survival probability.
As for the period 2003-2005 also showed inconsistent sequence as in the 2004-2006 period but in a different order.
The lowest survival probability was suitable with the estimation that belongs to firms which have the highest fuel cost share.
However, the highest survival probability did not belong to the firms with the lowest fuel cost share but firms with the share between 10%-20%.
In the long run, the lowest survival probability belonged to the firms which had the highest share of the fuel cost .
ore than 50%).
While the second and third rank of the lowest survival probability belonged to the firms which had fuel cost share 40% 50% and 30% -40% respectively.
However, for the following order was not in accordance with the order of the share of fuel Survival probability was the highest in the firm's share of the cost of fuel with 20% -30%.
While firms with the lowest fuel cost share had a fourth rank survival It can be concluded that, rising fuel prices in the long run affected the survival probability with the firm which had high fuel cost share.
While rising fuel prices had little impact on the firms which had less than 30% fuel cost share.
It means, in the long term, rising fuel prices began to feel an impact for firms with a high share of fuel costs.
CONCLUSION
The death risk of firms in the manufacturing sectors had no different for firms which were in the sub-sectors of the food industry or machinery industry.
Even for the entire sub-sector of the manufacturing industry, all firms faced the same probability to survive.
Similarly to the ownership of the firm, or a foreign country .
tate/enterprise.
were equally likely to survive in the market.
Table 6: Survival Probability of Manufacturing Firms Based on Fuel Cost Share
Time Period
Fuel Cost
Share
2002-2003 2003-2004 2003-2005 2004-2006 2005-2006 2002-2006
<10%
10%-20%
20%-30%
30%-40%
40%-50%
>50%
Total Notes: The calculation based on Table 4 and each non dummy variables average.
The numbers in parentheses are the survival probability rank from the lowest .
until the highest .
Fuel Price Increase and ManufacturingA (Atmadj.
Unlike foreign-owned and stateowned firms, national private firms had higher survival probability in the market than other firms.
More experience about how to manage the firm in the face of the business environment in Indonesia was believed to be an important factor why firms with national private ownership had higher probability to survive.
At first, productivity did not significantly affect the firm's first in survival.
However since 2004, productivity became an important factor in increasing the probability of survival for manufacturing firms.
With a more favorable economic conditions after 2004, productivity became an important factor if a firm wanted to increase the survival probability.
Among the various factors of firm
characteristics, size and age were dominant in increasing the probability to survive for a The larger the size and the older the REFERENCES