Baseline analysis of productivity changes with and without considering carbon dioxide emissions in the major manufacturing sector of Indonesia

Armundito, Erik; Kaneko, Shinji Article Baseline analysis of productivity changes with and without considering carbon dioxide emissions in the major manufacturing sector of Indonesia Journal of Economic Structures Provided in Cooperation with: Pan-Pacific Association of Input-Output Studies (PAPAIOS) Suggested Citation: Armundito, Erik; Kaneko, Shinji (2015) : Baseline analysis of productivity changes with and without considering carbon dioxide emissions in the major manufacturing sector of Indonesia, Journal of Economic Structures, ISSN 2193-2409, Springer, Heidelberg, Vol. 4, pp. 1-24, https://doi.org/10.1186/s40008-015-0018-3 This Version is available at: https://hdl.handle.net/10419/147205 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ RESEARCH Open Access Baseline analysis of productivity changes with and without considering carbon dioxide emissions in the major manufacturing sector of Indonesia Erik Armundito 1,2* and Shinji Kaneko 1 * Correspondence: [email protected] 1 Graduate School for International Development and Cooperation (IDEC), Hiroshima University, 1-5-1, Kagamiyama, Higashi-hiroshima-shi, Hiroshima-ken 739-8529, Japan 2 National Development Planning Agency (BAPPENAS), Jl. Taman Suropati No. 2, Jakarta Pusat, DKI Jakarta 10310, Indonesia Abstract This paper provides empirical evidence of changes in the productivities of manufacturing firms in Indonesia over time, in the form of total factor productivity (TFP), from 1990 to 2010 with and without considering carbon dioxide (CO 2 ) emissions. Employing cleaned and balanced panel datasets for four periods, 1990–1995, 1998–2000, 2003–2006, and 2008–2010, the analysis enables an evaluation of the impact of implemented policies or economic circumstances during each period. The Malmquist productivity index is employed to estimate TFP without CO 2 emissions over time, whereas the Malmquist-Luenberger productivity index is applied to estimate TFP with CO 2 emissions over time. Furthermore, the influence of energy factors on environmental productivity is also investigated. The results show that on average, TFP with CO 2 emissions over time has grown faster than TFP without CO 2 emissions, particularly for periods 1, 2, and 4. Technical progress is the basis of productivity growth after removing energy subsidies, and the change in environmental productivity is associated with the adjusted energy prices. Constructive policy designs can be derived from this paper that will enhance manufacturing sector performance after changes in the prices of oil commodities. Keywords: Total factor productivity; Directional distance function; Malmquist-Luenberger productivity index; Manufacturing sector JEL Classification: C61; D24; O47 1 Background The abundance of fossil energy resources as well as a large population has been the foundation of development in Indonesia. However, since 2004, Indonesia has become a net oil-importing country if we consider the trade balance of both crude oil and petroleum commodities. In addition, as of 2013, Indonesia ranked as the 11th largest CO 2 - emitting country after Canada [1]. As a growing and developing country in Asia with a relatively large but young demographic structure, Indonesia will not only confront domestic policy challenges but will also begin to draw international attention after China and India in seeking a future development pathway that is less fossil energy resource dependent and that creates more job opportunities. Although these challenges should be addressed by various sectors as declared by Indonesia’s master plan of 2011, the manufacturing sector is one of the most important © 2015 Armundito and Kaneko. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. Armundito and Kaneko Journal of Economic Structures (2015) 4:6 DOI 10.1186/s40008-015-0018-3 sectors due to its large potential for creating job opportunities. At the same time, there is concern regarding the increasing demand for energy generated by the economic development policy through further industrialization and development of the manufacturing sector. Currently, total final energy consumption (TFEC) in the manufacturing sector represents 27.4 % of the TFEC of Indonesia in 2011, and this share has been growing steadily over the last two decades [2]. As the international oil price has increased since 2000 and has remained high compared to prices in the 1990s, the government of Indonesia as a net oil-importing country started to gradually remove subsidies for energy commodities starting in 2005. Consequently, the domestic price of oil commodities in Indonesia has been rising since this time, which has caused a significant financial burden for the manufacturing industry. Although economic instruments implemented within climate change mitigation policies such as a carbon tax have not yet been implemented in Indonesia, the recent rising price of domestic oil commodities can be seen as a quasi-carbon tax instrument because it has similar consequences. With these as background and motivation, this paper reports empirical evidence of changes in the total factor productivity (TFP) of manufacturing firms in Indonesia over time from 1990 to 2010 with and without considering CO 2 emissions. The comparison of the TFP with and without considering CO 2 emissions across different sectors of the manufacturing industry enables us to identify firm reactions to changes in the prices of oil commodities. It should be noted that although historical data for manufacturing firms in Indonesia are available from the datasets of annual manufacturing surveys conducted by the Indonesian Statistics Agency (BPS) for medium- and large-sized firms that employ at least 20 workers, the datasets contain inaccurate, incomplete, and erroneous data. Therefore, despite the availability of large sets of data, to the best of our knowledge, empirical studies of Indonesian manufacturing firms are limited. To overcome this constraint, we first developed a cleaned panel dataset from the annual survey data of medium- and large-sized firms in the manufacturing sector of Indonesia between 1990 and 2010, which is used for the present analysis. Because the system of firm identity codes was changed between 2000 and 2001, it is impossible to construct continuous annual firm datasets between two periods, namely 1990–2000 and 2001–2010. In addition, we found that some of key variables such as capital stock and energy consumption, which are necessary for the present analysis, are completely missing in the survey data for 1996, 1997, 2001, 2002, and 2007. Therefore, the cleaned and balanced panel datasets are constructed for only four periods: 1990–1995, 1998– 2000, 2003–2006, and 2008–2010. For these periods, the paper provides empirical results from the baseline analysis for productivity measurements. The remainder of the paper is organized as follows. provides a brief description of the contextual background of the analysis and a literature review. Section 2 explains the methodological approach. Section 3 presents the empirical results and a discussion, followed by concluding remarks in Section 4. The state-led industrial policy strategy began in the 1970s in Indonesia during the Suharto regime, driven by a large windfall in government oil revenue from 1973 to 1980 through the development of state-owned firms. The government strictly protected the state-owned firms and other domestic producers from international competition by providing tariff and non-tariff barriers, raw material subsidies, and credit subsidies in addition Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 2 of 24 to maintaining undervalued exchange rates [3, 4]. Whereas the share of the manufacturing sector to GDP increased, empirical studies generally agreed that there was no gain and potentially negative TFP growth in the manufacturing sector in Indonesia during the period between 1975 and 1985 [5–7]. In 1986, the devaluation of the exchange rate of the rupiah against the US dollar triggered a shift in foreign direct investment (FDI) from the Asian newly industrialized economies to Indonesia, which promoted a labor-intensive manufacturing base in industries such as textiles, shoes, wood products, and processed food. As a result of an increasingly open economic policy, the share of manufacturing products has significantly expanded in terms of foreign exports since the middle of the 1980s. Furthermore, with the development of the machinery industry driven by FDI, the export of machinery products has also increased since the late 1980s and early 1990s. The shift to an FDI-led import substitution policy of industrialization has resulted in an increase in TFP for the manufacturing sector of Indonesia. Table 1 summarizes the results of the selected literature measuring the TFP growth of the Indonesian manufacturing sector using firm-level data for the period between 1970 and 2000. Although specific periods and numbers are not exactly the same and comparable, general shifts in TFP growth before and after the middle of the 1980s are commonly and consistently reported. Moreover, TFP growth seems to have continued until 1997, at which point the Asian economic crisis hit the Indonesian economy. Suharto’s relinquishing presidential office in 1998 evidences the seriousness of the adverse effects from the Asian financial crisis on the Indonesian economy, and this crisis also caused significant turbulence and confusion in measurements of TFP growth. Table 1 TFP growth measurements using firm-level data for the Indonesian manufacturing sector Authors Methods Periods Annual TFP growth (%) 1. Timmer [5] Growth accounting method 1975–1981 1.1 1982–1985 0.1 1986–1990 7.9 1991–1995 2.1 1975–1995 2.8 2. Aswicahyono and Hill [6] Growth accounting method 1976–1980 1.1 1981–1983 −4.9 1984–1988 5.5 1989–1993 6.0 1975–1993 2.7 3. Vial [7] Cobb-Douglas production function 1976–1980 1.5 1981–1983 −0.1 1984–1988 5.1 1989–1993 8.0 1976–1996 3.5 4. Ikhsan-Modjo [34] Stochastic production frontier 1988–1992 2.7 1993–1996 2.9 1997–2000 −0.6 1988–2000 1.6 TFP total factor productivity Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 3 of 24 As mentioned earlier, the datasets used in our analysis have several breaks, and the period before and after the Asian financial crisis is one of these breaks. Table 2 provides a summary of the key variables in the four analyzed periods to describe the contextual background of the present analysis. Period 1 from 1990 to 1995, which is the longest among the four analyzed periods in the paper, exhibited the highest average GDP growth rate at 7.9 %, and the growth rate of the manufacturing sector during this period was also the highest. Consequently, the share of the manufacturing sector to GDP increased from 21.6 to 24.5 %, and the share in total merchandise exports also increased from 35.5 to 50.6 %. Furthermore, the share of high-technology exports to manufactured exports expanded substantially from 1.6 to 7.3 %. In contrast, the growth rate of total final energy consumption (TFEC) for Indonesia and the growth rate of the manufacturing sector grew less quickly than that for production, resulting in an elasticity of TFEC to GDP of 0.57 and 0.75, respectively. Although net crude oil exports and the share of fuel exports to merchandise exports have been declining during the period, trade surpluses of more than 30 million TOE of crude oil were maintained. Overall, the last phase of the Suharto regime can be summarized as a time when the productivity and energy efficiency of the manufacturing sector was improved through an export-led industrialization policy. Period 2 from 1998 to 2000 is characterized as an immediate post-economic crisis period and marks the beginning of democratic reforms after the Suharto regime. Per capita GDP in constant US dollars at 2005 prices moved to an even lower range compared to 1995, and the average GDP growth rate was only 2.8 % during the period. However, the manufacturing sector performed relatively better despite the negative effects of the financial crisis. The share of the manufacturing sector to GDP slightly expanded from 26.0 to 27.1 % and that of exports to merchandise increased from 45.0 to 57.1 %. At the same time, the share of high-technology exports to manufactured exports also continued to increase from 10.4 to 16.4 %. However, energy consumption in Indonesia sharply increased during this time, and the elasticity of TFEC to GDP was 2.07, whereas the elasticity of the manufacturing sector was 1.25. Net crude oil exports started to decline from 27.3 to 17.4 million TOE, and the net export of oil products turned negative during this period. Period 3 between 2003 and 2006 covers a politically significant transitional moment when President Yudoyono became the first president of the country elected by a direct presidential election in 2004. Immediately after electing a new president, the Sumatra-Andaman earthquake and tsunami hit the country. The period first experienced a transition from positive to negative net oil exports considering both crude oil and oil commodities. Coincidentally, unprecedented and continuously soaring international oil prices finally forced the government of Indonesia to begin removing subsidies for oil commodities twice in 1 year in March and October 2005, doubling the prices for most oil commodities in the domestic market. Under these conditions, the manufacturing sector grew annually by 5.2 % on average, which was slightly lower than GDP growth. Meanwhile, the share of manufactured exports to merchandise exports dropped from 52.1 to 44.7 %, and the share of hightechnology exports to manufactured exports also shrank from 14.8 to 13.5 %. The energy intensityofthecountryasmeasuredbytheratioofTFECtoGDPgreatlyimproved,while the energy intensity of the manufacturing sector worsened. It is expected that manufacturing firms faced a significant increase in energy costs. Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 4 of 24 Table 2 Summary of key variables for four analytical periods Period 1 (6 years) Period 2 (3 years) Period 3 (4 years) Period 4 (3 years) Variables Unit Source 1990 1995 1998 2000 2003 2006 2008 2010 Per capita GDP USD at 2005 price a 840.2 1129.1 1057.1 1086.1 1180.5 1324.5 1451.6 1570.2 GDP growth rate % a 7.9 2.8 5.4 5.4 Growth rate of value added in manufacturing sector a 10.6 4.9 5.2 3.5 Share of manufacturing sector to GDP % a 21.6 24.5 26.0 27.1 27.3 27.2 26.1 25.2 Share of manufactures exports to merchandise exports % a 35.5 50.6 45.0 57.1 52.1 44.7 38.8 37.5 Share of high-technology exports to manufactured exports % a 1.6 7.3 10.4 16.4 14.8 13.5 10.9 9.8 Total final energy consumption (TFEC) 1000 TOE b 79,817 99,513 107,332 120,323 128,043 139,427 139,686 156,113 Growth rate of TFEC % b 4.5 5.9 2.9 5.7 Elasticity of total TFEC to GDP –a/b 0.57 2.07 0.53 1.05 Energy intensity (TFEC/GDP) TOE/USD at 2005 price a/b 531.8 454.1 500.2 530.2 497.2 462.3 410.8 413.1 TFEC in manufacturing sector 1000 TOE b 17,805 26,087 26,914 30,333 33,548 43,820 39,971 45,264 Growth rate of TFEC in manufacturing sector % b 7.9 6.2 9.3 6.4 Elasticity of TFEC to GDP in manufacturing sector –a/b 0.75 1.25 1.80 1.85 Energy intensity of manufacturing sector TOE/USD at 2005 price a/b 548.8 486.3 482.3 493.5 476.5 534.8 449.6 475.6 Net export of crude oil 1000 TOE b 32,328 30,744 27,349 17,390 7043 (2860) 1350 (2171) Net export of oil products 1000 TOE b 8054 2675 1529 (4181) (6896) (11,598) (15,860) (20,722) Share of fuel exports to merchandise exports % a 44.0 25.4 19.1 25.4 25.8 27.2 29.1 29.7 Note: a) World Development Indicators 2014, b) IEA Energy Balance Tables for Non-OECD Countries, 2013 GDP gross domestic product, TOE tons of oil equivalent, USD US dollar Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 5 of 24 Period 4 from 2008 to 2010 was in the middle of the 10-year presidency of Yudoyono and of the global financial crisis triggered by the subprime mortgage crisis and the bankruptcy of Lehman Brothers in 2008. The adverse shock caused in Indonesia by the global financial crisis was relatively small, and the average GDP growth rate in period 4 was maintained, staying as high as that of period 3. However, the growth rate of the manufacturing sector slowed and the shares of the manufacturing sector to GDP, manufactured products to merchandise exports, and high-technology products to manufactured products all shrank. At the same time, dependency on imported oil commodities increased remarkably, whereas the average growth rate of TFEC in the manufacturing sector was 6.4 %, which is much higher than production growth, resulting in an elasticity of 1.85. Amid such circumstances, the overall energy intensity of manufacturing firms did not improve. Further subsidy removal was implemented in 2008, and it is likely that the additional burden put a strain on manufacturing firms. A number of studies have attempted to analyze changes in productivity addressing multiple outputs, including both desirable and undesirable outputs. Data envelopment analysis (DEA) is one approach commonly employed to measure productive efficiency and is known as a non-parametric frontier approach [8]. DEA develops a nonparametric envelopment frontier encompassing all sample data as observed points lying on or below the frontier. The points on the production frontier are considered to be efficient decision-making units (DMUs), and the points below the production frontier are regarded as inefficient DMUs. The efficiency of each observation is measured by calculating the distance between the observed level of production and the production frontier as solutions of a linear programming problem. However, the DEA method does not evaluate the shift in the frontier over time; instead DEA only estimates the performance of DMUs in reference to the best practice frontier in a given year. The Malmquist productivity index is then introduced by adjusting the DEA application for multiyear observations alternately between tand t+ 1 to account for a shift in the frontier and to allow the measurement of changes in productive efficiency over time. The index measuring change in productive efficiency is then regarded as representing TFP growth, which can be further decomposed into efficiency change (catch-up) and technical progress (frontier-shift). Several ideas and methods have been proposed to incorporate undesirable outputs into DEA approaches while assuming asymmetrical treatments of disposability between desirable outputs and undesirable outputs when production possibilities are defined [9]. An efficiency improvement strategy for inefficient DMUs is developed by holding one or two inputs, desirable outputs or undesirable outputs constant. For example, input orientation refers to a strategy that considers how much inputs can be reduced while holding both desirable and undesirable outputs unchanged (i.e., [10]). Some others suggest a bads orientation strategy, which considers how bads can be reduced while holding inputs and desirable outputs unchanged (i.e., [11]). Tyteca [11] also considers the third strategy, which examines how much bads and inputs can be reduced while holding desirable outputs unchanged. The other recent strategy to incorporate undesirable outputs is to consider the proportional reduction in inputs and undesirable outputs and the proportional increase in desirable outputs [12]. Based on the earlier method of simultaneous change in desirable outputs and undesirable outputs following a hyperbolic function where fixed inputs are assumed, Chung et al. [13] proposed an Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 6 of 24 application of a directional distance function as well as a productivity index known as the Malmquist-Luenberger productivity index. The directional distance function (DDF) defines a strategy in which desirable and undesirable outputs are simultaneously changed. While the efficiency measurement with a DDF model in a single year is measured as the Luenberger productivity index, an alternate application of the DDF model between tand t+ 1 to measure the Luenberger productivity index in such a way that a Malmquist productivity index is constructed can generate the Malmquist-Luenberger productivity index. Therefore, the Malmquist-Luenberger productivity index can be further decomposed into efficiency change (catch-up) and technical progress (frontier-shift). Since Chung et al. [13] reported empirical results for Swedish firms in the paper and pulp industry using the directional distance function to develop the Malmquist- Luenberger productivity index, the Malmquist-Luenberger productivity index has been widely used in various studies at different levels from micro (firm-level data), to industry (sector-level data), to macro (province, national, and regional-level data) for evaluating productivity changes considering undesirable outputs. Färe et al. [14] employ the Malmquist-Luenberger productivity index to study the US manufacturing sector from 1974 to 1986 and observed that average annual productivity growth was 3.6 % when both desirable and undesirable outputs were considered, whereas it was 1.7 % when undesirable outputs were ignored. A comparable finding was presented by [15], who employed a similar approach to estimate productivity growth for six US chemical industries for the period 1988 to 1993; they concluded that environmental protection measures did not reduce productivity growth. Several other studies, including He et al. [16] and Piot-Lepetit and Moing [17] have focused on micro-level issues, while Kumar [18], Oh [19], and Zhang et al. [20] focused on macro-level issues using state- and regional-level data. Furthermore, the Malmquist-Luenberger productivity index has also been applied to industrial-level issues by Boyd et al. [21], Heng et al. [22], and Krautzberger and Wetzel [23]. TFP growth might be greater or smaller when undesirable outputs are considered compared to TFP measurement without considering undesirable outputs. Table 3 compiles several studies that analyze and examine TFP growth using the Malmquist- Luenberger productivity index with various topics, variables, and time periods. The results are discussed based on environmental regulation being the key determinant of the difference in productivity measurements with and without undesirable outputs. Chung et al. [13], Kumar [18], and He et al. [16] confirmed that TFP growth is higher when undesirable outputs are considered under stringent environmental regulations. However, these findings contrast with Zhang et al. [20], who analyzed TFP growth in China’s 30 provincial regions, implying that environmental regulations are not very stringent or not strictly enforced. Further indexing of TFP is proposed by Färe et al. [10] to elucidate the net contribution of environmental factors to productivity growth. Following this concept, Managi and Jena [24] estimate the ratio of TFP considering CO2 emissions to TFP without considering CO2 emissions as an environmental productivity measurement, which is referred to as the TFP environment. An increase in the TFP environment is considered to represent a positive achievement of proactive environmental measures or how well environmentally friendly technologies and managements are utilized [25]. Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 7 of 24 Table 3 Selected studies using the Malmquist-Luenberger productivity index to analyze TFP growth Authors Topics TFP growth Unit of analysis Period Without undesirable output (%) With undesirable output (%) 1. Chung et al. [13] Swedish pulp and paper industry: Firm 1986–1990 −0.3 3.9 Inputs: labor, wood fiber, energy, capital Desirable outputs: pulp Undesirable outputs: BOD, COD, and SS 2. Färe et al. [14] US State manufacturing air pollution emission: Industry 1974–1986 16.9 36.3 Inputs: employees, capital Desirable output: Gross State Product Undesirable outputs: SOx, CO 3. Kumar [18] 41 developed and developing countries: Country 1971–1992 −0.002 0.02 Inputs: labor, capital, energy consumption Desirable output: GDP Undesirable output: CO 2 4. Zhang et al. [20] China’s 30 provincial regions: Province 1989–2008 4.84 2.46 Inputs: labor, capital Desirable output: GDP Undesirable output: SO 2 5. He et al. [16] China’s iron and steel industry: Firm 2006–2008 19.2 19.8 Inputs: net fixed assets, employees, energy Desirable output: value added Undesirable output: waste water, waste gas, solid waste TFP total factor productivity, GDP gross domestic product, BOD biochemical oxygen demand, COD chemical oxygen demand, SS suspended solids, SOx sulphur oxide Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 8 of 24 Data on quantities of inputs, desirable outputs, and undesirable outputs are required to estimate productivity change over time using the Malmquist productivity index and the Malmquist-Luenberger productivity index. All firms are assumed to share the same production processes, characterized by the production of one desirable output and one undesirable output. Value added to manufacturing production and CO 2 emissions are considered to be the proxies for desirable (y) and undesirable (b) outputs, whereas capital (x1), labor wages (x2), and raw materials (x3) are considered as inputs. The value added vis measured as the difference between the total sales revenue of a firm and the total cost of components, materials, and services in millions of US dollars. Capital kis measured by the replacement value of fixed assets in thousands of US dollars. Labor wage lis measured as the total salary and other incentives for all workers, including production workers and other workers, in thousands of US dollars. Raw material mis measured as the total materials used to produce a unit of output in thousands of US dollars. Finally, both direct and indirect CO 2 emissions are measured as the most common type of gas emitted from the burning of fossil fuels used in manufacturing firms in tons CO 2 equivalent. Direct and indirect CO 2 emissions are calculated from fuel combustion in the manufacturing sector based on the Intergovernmental Panel on Climate Change (IPCC) guidelines [33]. Further, Indonesia’s currency rate has devalued since 1998 and a high inflation rate occurred in 1998, resulting in a monetary value for some variables in periods 3 and 4 that are smaller than the monetary value in periods 1 and 2. To measure the determinant of environmental productivity, three variables are included: the average domestic fuel price, the average electricity price, and energy dependency. The average domestic fuel price in constant prices (US dollars/TOE) is obtained as the ratio of total energy expenditure to total energy consumption. The average electricity price in constant prices (US dollars/TOE) is the ratio of total electricity expenditure to total energy consumption. Energy dependency (%) is defined as the ratio of total energy expenditures (US dollars) to total intermediate input expenditures (US dollars). The descriptive statistics for the variables are presented in Table 4. 3 Results and discussion Under two assumptions of the disposability of undesirable outputs, the Malmquist productivity index is applied to estimate TFP without CO 2 emissions over time and the Malmquist-Luenberger productivity index is employed to measure TFP with CO 2 emissions over time. A summary of the estimation results on an average annual basis for period 1 from 1990 to 1995 is presented in Table 5. Of the measurements without CO 2 emissions, the average productivity index score is 1.0014, implying that the annual TFP without CO 2 emissions over time for the manufacturing sector increases by 0.14 % over the entire period. This annual TFP score is obtained as the weighted mean of all sectors’TFP scores because the number of firms is different for each sector. On average, this growth is due to an increase in efficiency change of 0.97 % and a decrease in technological progress of 0.01 %. Based on the sector-by-sector analysis, considerable variation across sectors is observed. The sector that exhibits the highest productivity growth is motor vehicles, trailers, and semi-trailers (1.64 %), and the sector with the Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 15 of 24 Table 4 Descriptive statistics of the variables used for all periods Variable code Description Unit Mean Standard deviation Period of 1990–1995 kCapital Thousands of US dollar 450.80 1691.45 lLabor wage Thousands of US dollar 263.94 873.13 mRaw material Thousands of US dollar 607.63 2774.68 vValue added Thousands of US dollar 317.11 1370.26 CO 2 CO 2 emissions Tons CO 2 equivalent 584.88 2113.26 YEnvironmental productivity change over time % 1.15 2.72 Energydep Energy dependency % 2 0.02 Fuelprice Fuel price US dollar/tons of oil eq. 40.88 16.17 Elecprice Electricity price US dollar/tons of oil eq. 152.99 82.46 No. of observations 9336 Period of 1998–2000 kCapital Thousands of US dollar 352.88 1530.00 lLabor wage Thousands of US dollar 86.35 404.82 mRaw material Thousands of US dollar 596.34 3302.94 vValue added Thousands of US dollar 381.29 3211.77 CO 2 CO 2 emissions Tons CO 2 equivalent 666.49 2229.87 YEnvironmental productivity change over time % 1.10 1.10 Energydep Energy dependency % 5 0.06 Fuelprice Fuel price US dollar/tons of oil eq. 67.47 25.99 Elecprice Electricity price US dollar/tons of oil eq. 194.83 97.90 No. of observations 4668 Period of 2003–2006 kCapital Thousands of US dollar 127.14 408.26 lLabor wage Thousands of US dollar 86.27 216.78 mRaw material Thousands of US dollar 236.26 1489.42 vValue added Thousands of US dollar 119.65 583.46 CO 2 CO 2 emissions Tons CO 2 equivalent 282.02 1434.19 YEnvironmental productivity change over time % 1.154 1.04 Energydep Energy dependency % 8 0.08 Fuelprice Fuel price US dollar/tons of oil eq. 215.66 104.18 Elecprice Electricity price US dollar/tons of oil eq. 696.57 431.92 No. of observations 3296 Period of 2008–2010 kCapital Thousands of US dollar 120.00 441.86 lLabor wage Thousands of US dollar 84.76 253.55 mRaw material Thousands of US dollar 298.67 1924.40 vValue added Thousands of US dollar 119.50 385.95 CO 2 CO 2 emissions Tons CO 2 equivalent 243.63 1054.26 YEnvironmental productivity change over time % 1.08 0.56 Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 16 of 24 lowest growth is fabricated metal products and equipment (−0.07 %). Furthermore, for the measurement with CO 2 emissions, the weighted mean of the TFP scores is 1.0198, indicating that the annual TFP with CO 2 emissions over time for the manufacturing sector increases by 1.98 % over the entire period. This increasing growth is considerably higher than the growth of TFP without CO 2 emissions. This average TFP with CO 2 emissions over time is due to an increase in efficiency change (7.69 %) and technological progress (5.87 %). The sector that shows the highest productivity growth is other non-metallic mineral products (11.1 %) and the sector with the lowest growth is chemicals and chemical products (−2.7 %). At the same time, the average TFP environment score for period 1 is 1.0184, suggesting that environmental productivity increases by 1.84 % annually. The productivity measurement for period 2 from 1998 to 2000 is presented in Table 6. The average of the weighted means productivity index score is 0.9933, indicating that the annual TFP without CO 2 emissions over time for the manufacturing sector dropped by 0.67 % over the entire period. This change is triggered by an increase in efficiency change (1.58 %) and a decrease in technological progress (1.94 %). Based on the sector- Table 4 Descriptive statistics of the variables used for all periods (Continued) Energydep Energy dependency % 13 0.12 Fuelprice Fuel price US dollar/tons of oil eq. 518.55 186.15 Elecprice Electricity price US dollar/tons of oil eq. 1007.62 660.44 No. of observations 2472 Table 5 Average annual changes in productivity growth and its components for period 1 Sector TFP growth without CO 2 emissions TFP growth with CO 2 emissions TFP environment M EFFCH TECH ML EFFCH TECH Food products and beverages 1.0041 0.9990 1.0080 1.0303 1.1903 1.5534 1.0261 Tobacco 1.0102 1.0431 0.9722 1.0199 1.1650 0.9346 1.0096 Textiles 1.0073 1.0259 0.9902 1.0052 1.1220 1.0507 0.9980 Wearing apparel 0.9985 0.9886 1.0115 1.0174 1.0305 1.0115 1.0189 Tanning and dressing of leather 0.9907 1.0465 0.9520 0.9402 1.1152 0.8797 0.9491 Wood and products of wood and plaiting 1.0027 1.0118 0.9915 0.9783 1.0507 0.9713 0.9757 Paper and paper products 1.0094 1.0205 0.9947 0.9915 1.0964 0.9463 0.9823 Publishing, printing, and reproduction 0.9943 0.9931 1.0020 1.0154 1.1494 0.9776 1.0212 Chemicals and chemical products 1.0053 0.9982 1.0106 0.9730 0.9372 1.1142 0.9679 Rubber and plastics products 1.0050 0.9999 1.0099 0.9797 1.0146 1.0938 0.9748 Others non-metallic mineral products 0.9916 1.0070 0.9871 1.1110 1.1159 1.0953 1.1205 Basic metals 1.0318 1.0346 0.9980 1.0165 1.0182 1.0433 0.9852 Fabricated metal products and equipment 0.9903 1.0041 1.0031 0.9924 1.0795 1.0858 1.0021 Machinery and equipment 1.0017 1.0360 0.9739 1.0327 1.2134 0.9832 1.0310 Electrical machinery and apparatus 1.0070 0.9630 1.0548 1.0065 0.9444 1.1185 0.9995 Motor vehicle, trailers, and semi-trailers 1.0164 1.0005 1.0195 1.0105 0.9965 1.0517 0.9943 Other transport equipment 1.0101 1.0124 1.0113 1.0619 1.1018 1.1400 1.0513 Furniture and manufacturing 0.9923 0.9898 1.0074 0.9971 1.0436 1.0061 1.0049 Weighted mean 1.0014 1.0097 0.9999 1.0198 1.0769 1.0587 1.0184 TFP total factor productivity Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 17 of 24 by-sector analysis, the sector that shows the highest productivity growth is textiles (0.49 %), and the sector with the lowest productivity growth is paper and paper products (−3.04 %). For the measurement with CO 2 emissions, the weighted mean of the TFP score is 1.0652, implying that the annual TFP with CO 2 emissions over time increases by 6.52 % over the entire period. This increasing growth is significantly higher than the growth of TFP without CO 2 emissions. This remarkable growth is caused by an increase in efficiency (8.06 %) and technological progress (5.68 %). The sector that demonstrates the best-performance is rubber and plastics products, with productivity growth of 20.47 %, and the sector with the worst performance is paper and paper products, with a decrease in productivity of 2.22 %. Meanwhile, the TFP environment score for period 2 is 1.0724, suggesting that environmental productivity increases by 7.24 % annually. The productivity measurement for period 3 from 2003 to 2006 is presented in Table 7, and the average TFP without CO 2 emissions score is 1.0050. The average TFP score indicates that the annual TFP without CO 2 emissions increases by 0.50 % over the entire period. Increases in the efficiency change (0.49 %) and technological progress (1.55 %) are the sources of this TFP growth. Tobacco is the sector that shows the highest productivity growth (4.45 %), whereas textiles sector exhibits the lowest productivity growth (1.71 %). At the same time, the weighted mean score of TFP with CO 2 emissions is 1.0036, implying that the annual TFP with CO 2 emissions over time increased by 0.36 % over the entire period. The increase in efficiency change (7.18 %) and technological progress (7.57 %) are the engines of this TFP growth. The sector that shows the Table 6 Average annual changes in productivity growth and its components for period 2 Sector TFP growth without CO 2 emissions TFP growth with CO 2 emissions TFP environment M EFFCH TECH ML EFFCH TECH Food products and beverages 0.9933 1.0163 0.9774 1.0603 1.2147 0.8838 1.0674 Tobacco 0.9979 1.0054 0.9923 1.0129 1.0670 0.9754 1.0150 Textiles 1.0049 1.0105 0.9965 1.0448 1.1088 0.9967 1.0397 Wearing apparel 0.9985 1.0236 0.9797 1.1001 1.2215 0.9992 1.1017 Tanning and dressing of leather 1.0002 0.9941 1.0108 1.0559 0.9083 1.4272 1.0557 Wood and products of wood and plaiting 0.9953 0.9746 1.0243 0.9778 0.7279 1.5209 0.9824 Paper and paper products 0.9696 0.9944 0.9825 0.9286 0.8073 1.2034 0.9577 Publishing, printing, and reproduction 0.9754 1.0192 0.9641 0.9674 0.9221 1.0727 0.9918 Chemicals and chemical products 0.9947 1.0293 0.9703 1.1403 1.0646 1.0855 1.1464 Rubber and plastics products 0.9941 1.0109 0.9875 1.2047 1.1510 1.0871 1.2118 Others non-metallic mineral products 0.9952 1.0039 0.9933 1.1013 1.1688 0.9872 1.1066 Basic metals 0.9939 1.0757 0.9339 1.0869 1.4516 0.8474 1.0936 Fabricated metal products and equipment 0.9825 1.1075 0.8949 1.0471 1.5578 0.7498 1.0657 Machinery and equipment 0.9774 1.0396 0.9440 0.9954 1.1737 0.9150 1.0184 Electrical machinery and apparatus 0.9795 0.9774 1.0043 1.0224 0.9711 1.0791 1.0439 Motor vehicle, trailers, and semi-trailers 0.9903 0.9726 1.0199 1.0516 0.9376 1.1286 1.0619 Other transport equipment 1.0010 1.0067 0.9972 1.0212 0.9297 1.1017 1.0202 Furniture and manufacturing 0.9978 1.0220 0.9784 1.0091 1.0678 0.9622 1.0113 Weighted mean 0.9933 1.0158 0.9806 1.0652 1.0806 1.0568 1.0724 Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 18 of 24 highest productivity growth is wood and products of wood and plaiting (11.37 %), and the sector with the lowest growth is chemicals and chemical products (−19.67 %). On average, the TFP environment score for period 3 is 0.9986, suggested that environmental productivity decreased by 0.14 % annually. The productivity measurement for period 4 from 2008 to 2010 is shown in Table 8. The average of the weighted means productivity index score is 1.0311, implying that the annual TFP without CO 2 emissions over time increases by 3.11 % over the entire period. This growth is due to an increase in efficiency change of 0.33 % and in technological progress of 3.65 %. The sector with the highest productivity growth is electrical machinery and apparatus. (6.72 %), and the sector with the lowest growth is tanning and dressing of leather (0.06 %). Furthermore, the measurement of TFP with CO 2 emissions results in a weighted mean TFP score of 1.0523, suggesting that the annual TFP with CO 2 emissions increases by 5.23 % over the entire period. The growth is generated by the increase in efficiency change (1.77 %) and technological progress (11.83 %). The sector that shows the highest productivity growth is basic metals (14.73 %) and the sector that presents the lowest productivity growth is tobacco (0.89 %). The TFP environment score for period 4 is 1.0206, implying that environmental productivity increases by 2.06 % annually. The comparisons between the periods enable an evaluation of the impact of implemented policies or economic circumstances as contextual background of this analysis on manufacturing performance. The export-led industrialization policy implemented in period 1 not only resulted in the highest level for the GDP growth rate and the growth Table 7 Average annual changes in productivity growth and its components for period 3 Sector TFP growth without CO 2 emissions TFP growth with CO 2 emissions TFP environment M EFFCH TECH ML EFFCH TECH Food products and beverages 0.9991 0.9724 1.0286 1.0100 0.9201 1.1757 1.0109 Tobacco 1.0445 1.0883 0.9698 1.1102 1.7478 0.8695 1.0629 Textiles 0.9829 0.9917 0.9940 0.9698 0.9885 1.0951 0.9867 Wearing apparel 0.9956 1.0288 0.9701 1.0064 1.3685 0.9862 1.0109 Tanning and dressing of leather 1.0438 1.0262 1.0169 1.0460 1.0516 1.0048 1.0021 Wood and products of wood and plaiting 1.0098 0.9497 1.0813 1.1137 1.0586 1.2980 1.1029 Paper and paper products 1.0042 1.0073 0.9967 1.0074 1.0321 0.9822 1.0032 Publishing, printing, and reproduction 1.0042 1.0073 0.9967 1.0074 1.0321 0.9822 1.0032 Chemicals and chemical products 0.9906 0.9726 1.0226 0.8033 0.7201 1.2626 0.8109 Rubber and plastics products 1.0140 1.0428 0.9764 0.9827 1.4561 0.6780 0.9691 Others non-metallic mineral products 1.0001 0.9653 1.0385 0.9714 0.8091 1.2040 0.9712 Basic metals 1.0370 0.9432 1.1123 1.0719 0.9100 1.2206 1.0336 Fabricated metal products and equipment 1.0345 0.9963 1.0450 0.9685 0.8542 1.2306 0.9362 Machinery and equipment 1.0229 0.9871 1.0360 1.0435 0.9220 1.1443 1.0201 Electrical machinery and apparatus 1.0443 1.0774 0.9854 1.0548 1.2721 0.9612 1.0100 Motor vehicle, trailers, and semi-trailers 1.0280 1.0108 1.0176 1.0367 1.0575 1.0440 1.0084 Other transport equipment 1.0055 1.0374 0.9726 1.0418 1.0333 1.0084 1.0361 Furniture and manufacturing 1.0003 0.9836 1.0185 1.0160 1.0583 1.2153 1.0157 Weighted mean 1.0050 1.0049 1.0155 1.0036 1.0718 1.0757 0.9986 Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 19 of 24 rate of the manufacturing sector but also positively influenced the growth of TFP with CO 2 emissions. In this period, the growth of TFP with CO 2 emissions over time was tenfold that of the growth of the TFP without CO 2 emissions. The TFP environment over time in period 1 also shows positive moderate growth. Considering period 2 to represent the immediate post-economic crisis period and the beginning of democratic reforms, the growth of TFP with CO 2 emissions over time was the highest. During the same period, the decline of average GDP growth and the growth rate for the manufacturing sector from the previous period resulted in negative growth for TFP without CO 2 emissions. Meanwhile, the TFP environment over time in period 1 also exhibited the highest growth, even surpassing the TFP with CO 2 emissions over time. The remarkable achievement of environmental measures in this period is consistent with the notable increase in the share of manufactured exports to merchandise exports and the share of high-technology exports to manufactured exports. Period 3 is regarded as a period of politically significant transitional moments, and the government of Indonesia began to remove subsidies for oil commodities. During period 3, the growth of TFP with CO 2 emissions over time dropped sharply compared with the previous period. Several sectors, particularly for the high energy-intensive sectors, might be negatively affected by the increase in prices of oil commodities. The TFP with CO 2 emissions over time of the high energy-intensive sectors: food and beverages; textiles and its related industry; chemicals and chemical product; rubber and plastics product; others non-metallic mineral product; and fabricated metal product and equipment sectors present considerable decline in this period. Only basic metals sector that Table 8 Average annual changes in productivity growth and its components for period 4 Sector TFP growth without CO 2 emissions TFP growth with CO 2 emissions TFP environment M EFFCH TECH ML EFFCH TECH Food products and beverages 1.0167 0.9672 1.0524 1.0450 1.0338 1.1743 1.0279 Tobacco 1.0178 0.9742 1.0454 0.9911 0.9483 1.1034 0.9738 Textiles 1.0209 0.9432 1.0853 1.0482 0.8590 1.2842 1.0267 Wearing apparel 1.0257 1.0027 1.0283 1.0506 1.0228 1.0659 1.0243 Tanning and dressing of leather 1.0006 0.9825 1.0198 1.0433 1.0420 1.0203 1.0426 Wood and products of wood and plaiting 1.0106 1.0217 0.9903 1.0437 1.2793 0.8615 1.0327 Paper and paper products 1.0302 0.9352 1.1086 1.1013 0.9327 1.2866 1.0690 Publishing, printing, and reproduction 1.0368 0.9064 1.1480 1.1205 0.8171 1.4283 1.0807 Chemicals and chemical products 1.0622 1.0002 1.0675 1.0875 0.9524 1.1803 1.0238 Rubber and plastics products 1.0408 1.0099 1.0370 1.0543 1.0127 1.0979 1.0130 Others non-metallic mineral products 1.0345 1.0613 0.9794 1.0499 1.1142 0.9681 1.0148 Basic metals 1.0333 1.0542 0.9798 1.1473 1.0788 1.0463 1.1103 Fabricated metal products and equipment 1.0277 0.9560 1.0801 1.0571 0.7765 1.4569 1.0286 Machinery and equipment 1.0378 1.0756 0.9841 1.0484 1.1844 1.0512 1.0102 Electrical machinery and apparatus. 1.0672 1.0750 0.9991 1.0582 1.0682 1.0525 0.9916 Motor vehicle, trailers, and semi-trailers 1.0475 1.0374 1.0121 1.0634 1.0838 1.0027 1.0152 Other transport equipment 1.0349 1.0461 0.9904 1.0565 1.1154 0.9563 1.0209 Furniture and manufacturing 1.0583 1.0111 1.0492 1.0627 0.9977 1.0929 1.0042 Weighted mean 1.0311 1.0033 1.0365 1.0523 1.0177 1.1183 1.0206 Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 20 of 24 shows insignificant decrease. The effort in reducing CO 2 emissions seems to have further pressures due to the increase in prices of oil commodities. A similarly worsened performance was also experienced by the TFP environment over time as the growth dramatically declined to reach a negative level. In contrast, TFP without CO 2 emissions over time indicated positive growth. Almost all of the sectors show a notable increase of this TFP, except for textiles and wearing apparel sectors. Despite the increase in energy costs, the growth rate in the manufacturing sector increased from the previous period. Furthermore in period 4, when the global financial crisis took place, the growth of all TFP measurements over time demonstrated a remarkable increase compared with period 3. Compared to the previous three periods, the growth of TFP without CO 2 emissions over time was the highest in this period even though the level of growth was still lower than that for TFP without CO 2 emissions over time. At the same time, the TFP environment over time also grew at a level equal to that in period 1. The economic and political policies implemented by the government during this period were able to address the adverse effects of the global financial crisis in Indonesia, particularly in the manufacturing sector. Overall, it is observed that TFP with CO 2 emissions over time has grown faster than TFP without CO 2 emissions for periods 1, 2, and 4. The faster growth of TFP with CO 2 emissions over time is consistent with Domazlicky and Weber [15], Färe et al. [14], and Kumar [18], who suggested that when accounting for changes in pollution as an undesirable output, the average productivity growth is higher than the growth when ignoring pollution. TFP with CO 2 emissions over time lower than TFP without CO 2 emissions was observed in period 3. However, in general, the manufacturing sector showed the best performance in period 4, characterized by a positive growth level for all TFP measures, including a positive growth level of its components: efficiency change and technological progress. Efficiency change is the source of productivity growth in periods 1 and 2, whereas technical progress is the basis of productivity growth in periods 3 and 4. In particular, three outstanding sectors have been noted as the bestperforming sectors with the highest productivity growth for all periods: motor vehicles, trailers, and semi-trailers; electrical machinery and apparatus; and basic metals. The best-performing sectors might be interpreted as a positive response to changes in the prices of oil commodities by maintaining the positive growth of TFP with CO 2 emissions over time during all periods. Because there is no empirical evidence confirming that the increase in energy costs might directly affect manufacturing productivity, the relationship of energy factors and environmental productivity is analyzed. Estimation results evaluating the influence of energy factors on environmental productivity changes over time in Indonesia’s manufacturing sector are presented in Table 9. Fuel price is statistically significant during 2003–2006 and shows a negative relationship to changes in environmental productivity. When energy is still subsidized during the 1990–1994 and 1998–1999 periods,theenergydependencyofthemanufacturingsectorisrelativelysmall.Fuel and electricity prices do not play a significant role in enhancing environmental productivity. The energy policies implemented after 2003 aimed to remove subsidies for oil commodities and caused fuel and electricity prices to increase. As energy dependency also increased, energy prices started to negatively influence environmental Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 21 of 24 productivity improvements. Hence, the change of the environmental component in productivity measurements, in terms of the level of CO 2 emission reduction, is associated with adjusted energy prices. In addition, the coefficients of the sectoral dummy variables are positive and statistically significant for particular periods. As the metal and mineral group sector is the basis of the sectoral dummy, the chemical, rubber, and plastic group has a greater effect, at a 10 % significance level, in period 1. In period 2, the food, beverage, and tobacco group has a greater effect at a 1 % significance level and the wood, paper, printing, and furniture group has a greater effect at a 10 % significance level than the metal and mineral group. In period 3, the chemical, rubber, and plastic group has a greater effect at 1 % significance level than the metal and mineral group, whereas in period 4, the food, beverage, and tobacco group has a greater effect at a 10 % significance level and the textile and leather group has a greater effect at a 5 % significance level than the metal and mineral group. 4 Conclusions This paper provides a baseline analysis of TFP growth over time with and without considering CO 2 emissions from 1990 to 2000. Considering the current data problems and missing key variable data, cleaned and balanced panel datasets are constructed for only four periods: 1990–1995, 1998–2000, 2003–2006, and 2008–2010. The four periods of cleaned and balanced panel datasets enable an evaluation of the impact of implemented policies or economic circumstances during each period. An assumption is made that undesirable outputs are weakly disposable because Indonesia has not implemented carbon regulations. The Malmquist productivity index is employed to estimate TFP without CO 2 emissions over time, and the Malmquist-Luenberger productivity index is applied to estimate TFP with CO 2 emissions over time. The influence of energy factors on environmental productivity changes over time is also investigated. The main findings of this paper can be summarized as follows. First, on average, TFP with CO 2 emissions over time has grown faster than TFP without CO 2 emissions, Table 9 Factors associated with changes in environmental productivity Independent variables Periods 1990–1994 1998–2000 2003–2006 2008–2010 Energydep −0.1222 −0.3973 −0.3792 −0.1314 Fuelprice 0.0011 −0.0011 −0.0005 b −0.0000 Elecprice −0.0004 −0.0000 −0.0000 −0.0000 c dsec1 0.1256 −0.1876 a −0.1629 −0.0857 c dsec2 −0.0012 −0.0040 −0.1773 −0.1253 b dsec3 −0.1320 −0.1299 c −0.0391 −0.0349 dsec4 −0.2077 c −0.0842 −0.4547 a −0.0169 dsec5 −0.0154 −0.0428 −0.1086 −0.0408 Constant 1.0805 1.0757 1.0976 1.2024 Number of observations 7795 3118 2451 1634 a A variable is significant at a 1 % level of significance b A variable is significant at a 5 % level of significance c A variable is significant at a 10 % level of significance Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 22 of 24 particularly for periods 1, 2, and 4. Second, efficiency change is the source of productivity growth in periods 1 and 2, whereas technical progress is the basis of productivity growth in periods 3 and 4. Third, climate change mitigation policy might lead to the increase of additional burden for the high energy-intensive sectors. And the bestperforming sectors, based on their ability to maintain the positive growth of TFP with CO 2 emissions over time in response to changes in the prices of oil commodities during all periods are shown by (i) motor vehicle, trailers, and semi-trailers, (ii) electrical machinery and apparatus, and (iii) basic metals. Fourth, the change in environmental productivity is significantly associated with adjusted energy prices. Several constructive policy designs can be derived from these findings. The results suggest that CO 2 emissions as undesirable outputs can be considered when measuring the manufacturing sector’s productivity growth as a response to climate change mitigation policy. Valuable lessons learned from the best-performing manufacturing sectors can be applied to other manufacturing sectors responding to changes in the prices of oil commodities. At the same time, technological improvement is expected to be a major concern for manufacturing firms’long-term strategic planning after changes in the prices of oil commodities. Competing interests The authors declare that they have no competing interests. Acknowledgements The authors thank two anonymous referees for their very helpful and constructive comments on an earlier draft of this paper, as well as the financial aid received from the Japan International Cooperation Agency (Project of Capacity Development for Climate Change Strategy in Indonesia). The authors are also grateful to Fuji Xerox Co., Ltd., and the Setsutaro Kobayashi Memorial Fund for additional financial support. The usual disclaimer applies. Received: 31 May 2015 Accepted: 12 June 2015 References 1. Global Carbon Project (2014) The Global Carbon Project. http://www.globalcarbonproject.org. Accessed 14 Apr 2015 2. IEA. Energy balances of non-OECD countries 2013. Paris: International Energy Agency; 2013. 3. Goeltom M. Indonesia’s financial liberalization: an empirical analysis of 1981–88 panel data. Institute of Southeast Asian Studies, Singapore: ASEAN Economic Research Unit; 1995. 4. Pangestu M. Economic reform, deregulation, and privatization: the Indonesian experience. Jakarta: Centre for Strategic and International Studies; 1996. 5. Timmer MP. Indonesia’s ascent on the technology ladder: capital stock and total factor productivity in Indonesian manufacturing, 1975–95. Bull Indones Econ Stud. 1999;35:75–97. doi:10.1080/00074919912331337497. 6. Aswicahyono H, Hill H. “Perspiration”vs “inspiration”in Asian industrialisation: Indonesia before the crisis. J Dev Stud. 2002;38:138–63. doi:10.1080/00220380412331322381. 7. Vial V. New estimates of total factor productivity growth in Indonesian manufacturing. Bull Indones Econ Stud. 2006;42:357–69. doi:10.1080/00074910601053227. 8. Coelli TJ, Rao DSP. Total factor productivity growth in agriculture: a Malmquist index analysis of 93 countries, 1980-2000. Agric Econ. 2005;32:115–34. doi:10.1111/j.0169-5150.2004.00018.x. 9. Färe R, Grosskopf S, Lovell C a K, Pasurka C. Multilateral productivity comparisons when some outputs are undesirable: a nonparametric approach. Rev Econ Stat. 1989;71:90–8. doi:10.2307/1928055. 10. Färe R, Grosskopf S, Tyteca D. An activity analysis model of the environmental performance of firms—application to fossil-fuel-fired electric utilities. Ecol Econ. 1996;18:161–75. doi:10.1016/0921-8009(96)00019-5. 11. Tyteca D. Linear programming models for the measurement of environmental performance of firms - concepts and empirical results. J Product Anal. 1997;8:183–97. 12. Nakano M, Managi S. Waste generations and efficiency measures in Japan. Environ Econ Policy Stud. 2012;14:327–39. doi:10.1007/s10018-012-0038-2. 13. Chung YH, Färe R, Grosskopf S (1997) Productivity and undesirable outputs: a directional distance function approach. J Environ Manage 51:229–240. http://dx.doi.org/10.1006/jema.1997.0146 14. Färe R, Grosskopf S, Pasurka C a. Accounting for air pollution emissions in measures of state manufacturing productivity growth. J Reg Sci. 2001;41:381–409. doi:10.1111/0022-4146.00223. 15. Domazlicky BR, Weber WL. Does environmental protection lead to slower productivity growth in the chemical industry? Environ Resour Econ. 2004;28:301–24. doi:10.1023/B:EARE.0000031056.93333.3a. 16. He F, Zhang Q, Lei J, Fu W, Xu X. Energy efficiency and productivity change of China’s iron and steel industry: accounting for undesirable outputs. Energy Policy. 2013;54:204–13. doi:10.1016/j.enpol.2012.11.020. Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 23 of 24 17. Piot-Lepetit I, Moing ML. Productivity and environmental regulation: the effect of the nitrates directive in the french pig sector. Environ Resour Econ. 2007;38:433–46. doi:10.1007/s10640-007-9086-7. 18. Kumar S. Environmentally sensitive productivity growth: a global analysis using Malmquist-Luenberger index. Ecol Econ. 2006;56:280–93. doi:10.1016/j.ecolecon.2005.02.004. 19. Oh DH. A global Malmquist-Luenberger productivity index. J Product Anal. 2010;34:183–97. doi:10.1007/s11123- 010-0178-y. 20. Zhang C, Liu H, Bressers HT a, Buchanan KS. Productivity growth and environmental regulations - accounting for undesirable outputs: analysis of China’s thirty provincial regions using the Malmquist-Luenberger index. Ecol Econ. 2011;70:2369–79. doi:10.1016/j.ecolecon.2011.07.019. 21. Boyd GA, Tolley G, Pang J. ) Plant level productivity, efficiency, and environmental performance of the container glass industry. Environ Resour Econ. 2002;23:29–43. 22. Heng Y, Lim SH, Chi J. Toxic air pollutants and trucking productivity in the US. Transp Res Part D Transp Environ. 2012;17:309–16. doi:10.1016/j.trd.2012.01.001. 23. Krautzberger L, Wetzel H. Transport and CO2: productivity growth and carbon dioxide emissions in the european commercial transport industry. Environ Resour Econ. 2012;53:435–54. doi:10.1007/s10640-012-9569-z. 24. Managi S, Jena PR. Environmental productivity and Kuznets curve in India. Ecol Econ. 2008;65:432–40. doi:10.1016/j.ecolecon.2007.07.011. 25. Kumar S, Managi S. Environment and productivities in developed and developing countries: the case of carbon dioxide and sulfur dioxide. J Environ Manage. 2010;91:1580–92. doi:10.1016/j.jenvman.2010.03.003. 26. Färe R, Grosskopf S, Lovell C (1994) Production frontiers. Cambridge: Cambridge University Press 27. Mandal S, Madheswaran S. Environmental efficiency of the Indian cement industry: an interstate analysis. Energy Policy. 2010;38:1108–18. doi:10.1016/j.enpol.2009.10.063. 28. Hamilton JD. A neoclassical model of unemployment and the business cycle. J Polit Econ. 1988;96:593. doi:10.1086/261553. 29. Hooker MA. Are oil shocks inflationary? Asymmetric and nonlinear specifications versus changes in regime. J Money Credit Bank. 2002;34:540–61. 30. Hamilton JD (2009) Causes and Consequences of the Oil Shock of 2007-08. National Bureau of Economic Research. Cambridge, MA. http://www.nber.org/papers/w15002.pdf 31. Aye GC, Dadam V, Gupta R, Mamba B. Oil price uncertainty and manufacturing production. Energy Econ. 2014;43:41–7. doi:10.1016/j.eneco.2014.02.004. 32. Armundito E, Kaneko S (2014) Note on Data Cleaning and Panel Data Development of Indonesian Manufacturing Survey Data (IDEC DP2 Series No. 4-9). Hiroshima University, Graduate School for International Development and Cooperation (IDEC) 33. Eggleston H, Buendia L, Miwa K, Ngara T, Tanabe K. IPCC guidelines for national greenhouse gas inventories. Hayama, Japan: Institute for Global Environmental Strategies; 2006. 34. Ikhsan-Modjo M (2006) Total factor productivity in Indonesian manufacturing: a stochastic frontier approach. Monash University (ABERU Discussion Paper 28). Submit your manuscript to a journal and benefi t from: 7 Convenient online submission 7 Rigorous peer review 7 Immediate publication on acceptance 7 Open access: articles freely available online 7 High visibility within the fi eld 7 Retaining the copyright to your article Submit your next manuscript at 7 springeropen.com Armundito and Kaneko Journal of Economic Structures (2015) 4:6 Page 24 of 24