Structural Change and Regional Economic Growth in Indonesia
Summary (4 min read)
- Structural change is an important determinant of economic growth.
- Empirically, the relationship between structural change and economic growth either in a regional or at a national level is rather inconclusive.
- Studies at the subnational level have also been conducted although mainly on China and India.
2.1. Measurement of Structural Change
- The measurements of structural change are usually calculated in terms of employment shares or value added shares.
- B. Norm Absolute Value (NAV) index NAV index is similar to SC index, but uses employment share rather than value added share, and is calculated as follows: 𝑁𝐴𝑉 = 12 ∑ |𝑆𝑖𝑇 − 𝑆𝑖0|𝑛𝑖=1 (2) where 𝑆𝑖𝑇 and 𝑆𝑖0 denote the employment share of sector 𝑖 at time T and 0, respectively.
- Again, this index only measures the shifting of employment, with no direct link between employment shifting and productivity, and does not make a distinction as to whether structural change experienced by a sector is productivity-enhancing or decreasing.
- The method decomposes the sectoral contribution to overall labor productivity growth into three terms as follows: ∆𝑃𝑃0 = ∑ 𝑆𝑖0∆𝑃𝑖𝑃0𝑛𝑖=1 + ∑ 𝑃𝑖0∆𝑆𝑖𝑃0𝑛𝑖=1 + ∑ ∆𝑆𝑖∆𝑃𝑖𝑃0𝑛𝑖=1 (3) where 𝑛, 𝑆𝑖𝑇 and 𝑆𝑖0 are the same variables as in the NAV index.
2.2. Regional Growth Model
- Silva and Teixeira (2008) classify the studies on structural change into 11 main topics, where this study focuses on the topics of convergence and growth as well as regional and urban economics.
- This study focuses on the direction from structural change to economic growth because it is interested in regional growth determinants.
- The authors also use a set of other control variables for excluded instruments (Panel B Table 1), which are the length of roads , the importance of mining , government spending for capital expenses , commercial banking loans and foreign direct investments (FDI).
- Collapsing technique and only use one lag to reduce the number of instruments used.
- GMM estimators are also used for annual data; however, the authors utilize the level GMM estimators instead of the two-steps system GMM system.
- In addition to the national level, this study calculated the four structural change measurements above for the 30 sub-nationals in Indonesia.
- It is worth noting that some of the provinces have experienced a fragmentation, such as Riau (some areas became Riau Islands), East Kalimantan (North Kalimantan), South Sulawesi (West Sulawesi) and Papua (West Papua).
- The data used for the calculation is collected from CEIC which is mainly derived from Badan Pusat Statistik (Indonesia Statistics) covering the 2005–2018 period (Panel A Table 1).5 INSERT TABLE 1 ABOUT HERE.
- Indonesia’s gradual structural transformation from a traditional agricultural-driven economy into a manufacturing and services-driven economy helped to boost Indonesia’s income per capita and to raise the nation’s status from an underdeveloped to a developing country by the late 1980s.
- The 1997–98 Asian financial crisis ended the episode of exponential growth abruptly, and Indonesia has not fully recovered from this crisis (Basri et al. 2016).
- Judging from the pace of economic transformation and industrialization, there is a clear disproportion among regions due to a poor national logistics system (Axelsson & Palacio 2018).
- In terms of the regional gross domestic product (RGDP)’s contribution to the total GDP.
- Griffith (1983) reports that the decision to define the boundary affects spatial distribution identification and statistical parameter estimation.
- Over the period 2005-2018, the progress of Indonesia’s structural transformation into a services-driven one has been slowing.
- The fragmentation of administrative entities at the sub-national level has been mirrored by a boom in the number of public-service jobs, at around 17.5 public servants per 1000 population (Vujanovic 2017).
- This indicates that workers tend to shift from productivityimproving sectors to productivity-declining sectors.
- The sectors experiencing both an increase in labor share and an increase in productivity are Manufacture, Construction, Trade, and Government.
- Agriculture, Manufacture, Utilities, Construction, Trade, and Transport, while in the period 2015–2018 the number of sectors has been reduced to five, i.e.
- Figure 2 illustrates the distribution of regional structural change in Indonesia.
- Panel A and Panel B map the effective structural change index and shift-share method (real labor productivity) 7 across provinces over the period 2005–5018, respectively.
- Almost all major provinces in Java Islands only grew in the range of the national average.
- While the provinces with the fastest growth and highest structural change index are mostly small provinces such as North Maluku, Maluku, Gorontalo, and Central Kalimantan.
INSERT FIGURE 1 ABOUT HERE
- Regionally, it is clear that the provinces in Sulawesi Island relatively have higher structural change measurements than those in Java Island,8 in particular in terms of NAV and ESC.
- This means that reallocation of employment to more productive sectors is happening more in Sulawesi Island than in Java Island.
- The share of the manufacturing sector in Riau, DKI Jakarta, Bali, West Nusa Tenggara, and East Kalimantan has been decreased.
- 8 Sulawesi Island, also known as Celebes, consists of five provinces: North Sulawesi, Central Sulawesi, South Sulawesi, South East Sulawesi, and Gorontalo.
- In terms of the within effect, the largest improvement in productivity was experienced by Riau in the period 2005–2018, while Aceh and East Kalimantan experienced a consistent decline in the same sector productivity in the four different periods of study.
3.2. Descriptive Statistics
- Table 3 and Table 4 present descriptive statistics for the variables of interest.
- Table 3 summaries the main economic data of Indonesia divided into 30 provinces.
- GDP from the poorest province is Rp 18 million per capita, lagging very far behind the richest province with Rp 248 million per capita.
- Most provinces experienced an inline trend with the national average, which increased in the 2010– 2014 period but declined in the 2015–2018 period.
- The commodity boom in the era of 2010–2014 plays an important factor here not only for natural resources-rich provinces but also by other regions that do not rely on natural resources.
INSERT TABLE 5 ABOUT HERE
- Table 5 shows the strength of the relationship among variables.
- Mostly there is a weak negative correlation between structural change measures and PRGDP, except for Static.
- In terms of cross-section dependence in macro panel data, Pesaran CD cross-sectional independent tests show that there is an interlinkage among provinces that may arise from globally common shocks as the result of local spillover effects between provinces.
- There may be a positive relationship between regional growth and the NAV index, which is similar to Hill et al. (2008) that also shows a weak relationship between structural change and regional growth.
- The next sub-section explains the formal examination of the relationships.
3.3. The Five-Year Average Data
- Table 5 presents the estimates of the traditional growth model using five-year average data.
- Model 1 to Model 4 presents the estimates for SC, NAV, ESC, and SS, respectively.
- The authors find that the coefficient estimate of SC is statistically insignificant, while other structural change measures are statistically significant.
- This means that the movement of labor across sectors may hamper economic growth if the movement does lead to higher productivity.
- The coefficient of Static is insignificant, even though positive.
3.4. The Annual Data
- In estimating the dynamic model using annual data, the authors firstly employ two panel unit root tests: the Im-Pesaran-Shin (IPS W-t-bar) test (Im et al. 2003) and the Pesaran's simple panel unit root test in the presence of the cross-section dependence (CADF Z-t-bar) test (Pesaran 2007).
- The test results generally show that (1) all variables contain a unit root except for structural change and government expenditure variables, and (2) there is no cointegration between regional economic growth and structural change measures.
- To be consistent with the interpretation of change, the authors estimate GMM by using first differences of all examined variables.
- For annual data, it seems that dynamic structural effects have more impact on growth than static structural effects and within-sector productivity improvement.
- This may due because the movement of labor occurs gradually.
- Recent studies from Hill et al. (2008) and Axelsson and Palacio (2018) have examined the pattern of structural change from Indonesia’s subnational perspective.
- The outperformer regions such as provinces in Sulawesi (except Gorontalo), Central Java, and East Java are characterized by their productivity that grows far above the national average, mainly driven by the growth of new industries.
- Meanwhile, the large gaps of share employment and value added in agriculture lead to a surplus of labor that has been unable to be absorbed by other more productive sectors (Axelsson & Palacio 2018).
- The speed of structural change in this region needs to be interpreted with great caution, given the low level of productivity and small economic share of the provinces to the national economy.
- In terms of the relationship between structural change and economic growth, this study has confirmed that there is a positive relationship between structural change and regional economic growth, in particular the shift-share method indicators.
- This paper has examined the dynamics of structural change and investigated the relationship between structural change and regional economic growth in Indonesia.
- By calculating four measures of structural change, namely structural change index, norm absolute value index, shift-share method, and effective structural change index, this study finds that structural change has occurred across 30 provinces over the period 2005–2018 toward an agricultural-services transition.
- The provinces in Sulawesi outperformed other regions.
- Even though the role of manufacturing needs to be improved as the engine of growth, industrialization policies must be based on the characteristics of each province.
- Structural change matters for growth only if there is an increase in productivity, not just the movement of labor across sectors.
Did you find this useful? Give us your feedback
Related Papers (5)
Frequently Asked Questions (15)
Q1. What have the authors contributed in "Structural change and regional economic growth in indonesia" ?
This paper investigates the relationship between structural change and regional economic growth in Indonesia. The authors show that the structural change has occurred across provinces, even though it is slowing, towards an agricultural-services transition. By employing dynamic panel data models, this study shows that structural change is a significant determinant of growth.
Q2. What is the main reason behind the failure of labor to more productive sectors?
Thedecentralization and proliferation of a district may be the main reason behind the failure ofreallocation of labor to more productive sectors due to diminishing marginal return in thegovernment sector.
Q3. What are the other important factors that may disrupt the structural change contribution to economic growth?
The other important factors that may disrupt thestructural change contribution to economic growth are the non-economic factors such as socialconflict and natural disaster (Rao & Vidyattama 2017; Heger & Neumayer 2019).
Q4. What tests are used to test the validity of the models?
Twospecification tests, i.e. the AR(2) test in first differences and the Hansen (1982) test concerning thejoint validity of the instruments, suggest that their models are acceptable.
Q5. How many public servants were transferred to the regions?
Mahi (2016) finds that over 2 million public servants or almost two-thirds of the centralgovernment workforce were transferred to the regions.
Q6. What are the main sectors that have seen the largest increase in productivity?
Manufacture, Utilities, Construction, Trade, and Transport, while in the period 2015–2018 the number of sectors has been reduced to five, i.e.
Q7. How can growth happen if there is an improvement in productivity within sectors?
Growth can happen if there is an improvement in productivity within sectors as well as by shifting to other sectorswith better productivity.
Q8. What are the main factors that may hamper the benefit of structural change?
The cost of transition, consisting of structural unemployment and social costs, on the otherhand, may hamper the benefit of structural change on economic growth (GHK 2011).
Q9. What is the only province that experienced the movement of labor to more productive sectors?
The only province that experienced the movementof labor to more productive sectors is Central Sulawesi, shown by a positive dynamic structural effect.
Q10. What is the main reason why the transition is happening in developing countries?
The phenomena of agriculture-services transition are commonly happening in developing countries as discussed by Chenery et al. (1986).
Q11. What is the effectivity of structural change in Maluku?
Structural changeoccurring in Maluku is more effective than other provinces shown by the higher ESC values andan increase in the number of sectors experiencing higher productivity.
Q12. What is the main reason why the agricultural sector is unable to absorb more labor?
the large gaps of share employment and value added in agriculture lead to a surplusof labor that has been unable to be absorbed by other more productive sectors (Axelsson &Palacio 2018).
Q13. What is the definition of structural change?
Krüger (2008), for instance, borrows the definition used by Erich Streissler, i.e. “long-term changes in the composition of economic aggregates,” while GHK (2011) defines structural change as “a dynamic and turbulent process associated with very substantial changes of growth and contraction at the sectoral1 O’Rourke and Williamson (2017) provide a comprehensive review of industrialization; while Assunção et al. (2015) show that the convergence depends on country-specific characteristics, such as political institutions, trade openness, and education.
Q14. What is the main difference between the shift-share methods and static structural effects?
In particular, the shift-share methodshows that for annual data dynamic structural effects have more impact on growth than staticstructural effects and within-sector productivity improvement.
Q15. What are the sectors that have the largest increase in productivity?
The sectors experiencing both an increase inlabor share and an increase in productivity are Manufacture, Construction, Trade, andGovernment.